Skip to main content

Full text of "House Documents of the House of Representatives 2nd session of the 58th congress 1903-1904"

See other formats





■•■ii mill lii 

ii ■ 



PROVO, UTAH 84602 

Si 2.1:901/1 

U.S. Ethnology Bureau reports. 

1 2 r 1 



'• ;5S 


Digitized by the Internet Archive 

in 2013 

58th Congress, ) HOUSE OF REPRESENTATIVES, j Document 

2d Session. f ( No. 041. 














Smithsonian Institution, 
Bureau of American Ethnology, 

Washington, D. C, July 1, 1899. 

Sir: I have the honor to submit, as Acting Director, 
the Twenty -second Annual Report of the Bureau of 
American Ethnology. 

The preliminary portion consists of an account of the 
operations of the Bureau during the fiscal year; the 
remainder comprises two memoirs, prepared by collab- 
orators, which illustrate the methods and results of the 
Bureau's work. 

Allow me to express my appreciation of your constant 
aid and your support in the work under my charge. 
I am, with respect, your obedient servant, 

W J McGee, 

Acting Director. 

Honorable S. P. L angle y, 

Secretary of the Smithsonian Institution. 




Report of the Director 


Introduction ix 

Field research and exploration x 

Office research xv 

Work in esthetology xv 

Work in technology xix 

Work in sociology xxiv 

Work in philology xxx 

Work in sophiology xxxiii 

Work in descriptive ethnology xxxvn 

Publication xxxvin 

Collections xxxvin 

Property xl 

Necrology — Frank Frederick Hilder xl 

Financial statement xlii 

Accompanying papers xliii 

Accompanying Papers 

Two summers' work in pueblo ruins, by Jesse Walter Fewkes (plates i-lxx, 
figures 1-122) 1 

Mayan calendar systems, II, by Cyrus Thomas (plates lxxi-lxxxii, figures 

123-170 197 


Accompanying Papers (Continued) 

The Hako, a Pawnee ceremony, by Alice C. Fletcher (plates lxxxiii-xci, fig- 
ures 171-181) 1 






J. W. Powell, Director 


Ethnologic researches have been conducted during the 
fiscal year ending June 30, 1901, in accordance with the act 
of Congress making provision "for continuing researches 
relating to the American Indians under the direction of 
the Smithsonian Institution," approved June 6, 1900, and 
with the formal plan submitted June 9, 1900, and approved 
by the Secretary June 19, 1900. 

The field operations of the regular corps extended into 
Arizona, British Columbia, California, Lower California 
(Mexico), Maine, New Mexico, New York, North Caro- 
lina, Ontario, Sonora (Mexico) , Virginia, and Wisconsin; 
while special work has been carried forward by agents or 
temporary collaborators in several additional states, ter- 
ritories, and provinces. The office work has comprised 
the preparation and study of material from most of the 
states and territories, as well as from various other parts 
of the American hemisphere. 

The researches have been carried forward in accord- 
ance with an ethnic system based chiefly on the work of 
the Bureau, though partly on the observations and deter- 
minations of other scientific investigators in this and 
other countries. 

The ethnic system developed and adopted in the Bureau 
is based primarily on the human activities — that is, on what 
men do and think — rather than on mere physical features. 



On this basis, the habits and customs of the aborigines 
receive first attention ; and the tribesmen are classed by 
their languages and dialects, by their forms of social 
organization, by their systems of belief and opinion, and 
by their arts and industries: so that the classification 
affords a means of measuring the susceptibility of the 
various tribes to civilization, to education, and to arrange- 
ment on reservations in harmonious groups. The classi- 
fication is thus essentially practical. 

The practical tribal classification rests on a definition 
of the activities discovered among the aborigines and 
other peoples largely during the past quarter -century. 
The primary activities thus discovered are esthetic ; and 
intimately connected with these are the industrial activi- 
ties involved in maintenance and welfare. Equally im- 
portant are the social activities shaping the collective 
existence of families, clans, tribes, and confederacies; 
and the relations are regulated by linguistic activities, 
which are highly important and indeed fundamental. 
Coordinate with these activities of arts and industries, 
laws and languages, are the activities connected with 
opinion, belief, philosophy — the sophic- activities. On 
weighing all the factors it has been found that the most 
convenient classification of tribes is that based primarily 
on language, as explained in previous reports; and this 
mode of defining the Indian tribes, first proposed by Gal- 
latin and adopted by the Bureau on its institution, has 
now come into general use. 


Throughout the first quarter of the year the Director 
was in Maine, reviewing observations on shell mounds 
and village sites in connection with the researches in 
classification noted in other paragraphs; and the work 
was resumed early in June. But limited collections were 
made, though the observations and notes on the numer- 
ous survivors of the Abnaki Indians proved of much 
interest and value. 


An extended exploratory trip was made during the 
autumn of 1900 by Mr McGee. Early in October he 
proceeded to the field for the purpose of completing 
researches relating to the aborigines of the Serian stock 
and at the same time carrying forward studies of neigh- 
boring tribes. A party was organized at Phoenix, Ari- 
zona, and moved southwestward to Gila bend and thence 
southward to the international frontier at Santo Domingo. 
Here the outfit was admitted to Mexican territory through 
the courtesy of Sefior Don Fernando Leal, at the obliging 
instance of Sefior Don Manuel de Aspiroz, the ambassa- 
dor from Mexico to the United States. In this vicinity 
are several settlements of Papago Indians, including some 
of the Arenefios of early literature and local tradition, 
and the opportunities for study were seized. From Santo 
Domingo the party proceeded southward to Caborca and 
thence westward to the coast of Gulf of California, where 
the Tepoka Indians (collinguals of the Seri) were reported 
to live so late as 1894, subsisting on sea food and finding 
potable water in the lagoons and sand beds at the em- 
bouchure of the sand wash variously called Magdalena, 
Santa Ana, Altar, Asuncion, and San Ignacio. On reach- 
ing the coast the leader was disappointed to find the 
tribal remnant entirely gone — probably through extinc- 
tion, possibly through migration down the coast to Seri- 
land. Traces of the Tepoka habitations still remained, 
together with shell accumulations and minor relics, cor- 
roborating the reports concerning the tribe current at 
Caborca in 1894; and the visit served also to clear up 
doubtful points connected with the geography and history 
of the region. 

Failing thus to attain the primary object of the expedi- 
tion, Mr McGee determined to visit the territory of the 
little -known Cocopa Indians, reputed to live about the 
head of the gulf, and to this end endeavored to follow 
the coast northward to the mouth of the Colorado. Find- 
ing this entirely impracticable, he returned by a new 
route to Santo Domingo, collecting useful data concerning 
the Papago Indians on the way ; and from Santo Domingo 


he proceeded west -northwestward over the old Yuma 
trail (including a stretch of 90 miles now without water) 
to Yuma, and thence southward to the Cocopa country. 
Here valuable collections, notes, and photographs were 
obtained; and after some weeks the party returned via 
Yuma and the Gila and Salado valleys to Phoenix, dis- 
banding there on December 20. The party comprised 
Mr W J McGee, Ethnologist in Charge, as leader; Mr 
DeLancey Gill, artist; Professor B-. H. Forbes, of the 
territorial university of Arizona (during part of the 
trip) ; Senor Aurelio Mata, a Mexican customs officer 
sent from the custom-house at Nogales to facilitate the 
crossing at the international boundary; John J. Carroll, 
of Tempe, teamster; Jim Moberly, of Tempe, packer; 
Hugh Norris, of Tucson, Papago interpreter, and Ramon 
Zapeda, of Tucson, Mexican interpreter. The Bureau 
was placed under great obligations for free entry of the 
outfit to the government of the neighboring republic 
through the officers already named, as well as through 
Senor Don Eduardo J. Andrade, of Yuma, custodian of 
the Andrade grant, covering the territory occupied by 
the Cocopa Indians. 

On August 11 Mr James Mooney proceeded to the old 
Cherokee country in western North Carolina and adjacent 
territory for the purpose of collecting additional data 
required for the completion of his series of papers on the 
Cherokee Indians, and his field operations continued with 
success until early December. On April 25 he made a 
reconnoissance trip through eastern North Carolina and 
Virginia for the purpose of locating remnants of aborigi- 
nal tribes still surviving in the wooded and nearly inac- 
cessible districts of that region; he revisited the Pamun- 
key tribe and discovered considerable remnants of the 
Chickahominy, Mattaponi, and Nansemond tribes. 

On his appointment as Assistant Ethnologist (Septem- 
ber 1) , Mr John R. S wanton proceeded to British Colum- 
bia to undertake researches among several northwestern 
tribes. His work proceeded successfully up to the end of 
the fiscal year, when he was still in the field. 


On October 1 Mr J. N. B. Hewitt repaired to the region 
occupied by the survivors of the Iroquoian tribes in north- 
western New York and neighboring portions of Canada, 
where he began the collection and verification of traditions 
and cosmogonic legends, and his work continued until 
about the middle of February, when he returned to the 
office with valuable collections and records. 

On April 15 Dr Frank Russell .was appointed as 
Ethnologist, and was assigned to duty in Arizona; he 
immediately proceeded to the field and began an extended 
reconnaissance of the southern and central portions of 
the Territory. Outfitting with a team at Tucson, he 
passed around the northern end of Santa Catalina moun- 
tains and up San Pedro river (visiting the caves and 
pictographs of the Santa Catalina range and the cliff 
houses of the Galiuro range on the way) to Nugent pass, 
where he entered Aravaipa valley. Here he found an 
interesting group of cliff houses. Thence he proceeded 
by way of Eagle pass to Gila valley, where interesting 
archeologic observations were made. Pushing on south- 
ward he traversed the eastern slopes of Chiricahua moun- 
tains and the western slopes of Swisshelm mountains, 
and examined the easterly canyons of Huachuca moun- 
tains. Next he traversed portions of the Babacomori, 
Sonoyta, and San Rafael valleys about .the Mexican 
boundary; thence he returned by new routes to Santa 
Catalina mountains and Tucson, arriving about the end 
of May. In the course of the trip he discovered various 
ruins hitherto unknown, some of new types. Several of 
the ruins were surveyed, and limited collections were 
made. On June 11 he proceeded northward from Tucson, 
crossing the Gila near Florence, skirting the base of 
Superstition mountains, and traversing Tonto valley; a 
number of cliff houses and other ruins were discovered 
during the journey, which was not completed at the end 
of the fiscal year. 

In June an arrangement was effected with Mr O. P. 
Phillips and the Armat Motion -Picture Company, under 
which Mr Phillips proceeded to New Mexico and Arizona 

22 ETH— 04 II 


for the purpose of making motion pictures representing 
the industries, amusements, and ceremonies of the 
Pueblo and other tribes, it being anticipated that such 
pictures would prove of especial service for purposes of 
immediate research as well as for permanent record. 
The preliminary reports indicate that the work has been 
successfully initiated. 

Throughout the fiscal year Dr Willis E . Everett remained 
in Alaska, pursuing his vocation of mining engineer, 
but availing himself of opportunities for observing the 
native tribes and recording their languages and other 
acti vital characteristics. Several reports indicating prog- 
ress in the collection of such material were received in 
the course of the year. 

Dr Robert Stein, who spent the winter of 1899-1900 in 
Elsmereland, primarily for purposes of geographic explo- 
ration, but incidentally to make search for traces of abo- 
riginal occupancy in the interests of the Bureau, reported 
via Dundee, through the courtesy of 'masters of whaling 
vessels, late in the summer of 1900. He found no traces 
of Eskimo or other settlements in the territory traversed 
by him, comprising the eastern coast of Elsmereland, and 
his negative evidence is of service in investigations relat- 
ing to the distribution and migrations of the Eskimo. 
At the time of the last report he was preparing to cross 
Baffin bay to Upernivik, on the western coast of Green- 
land, with the expectation of extending his previous 
observations on prehistoric Eskimo settlements along the 
unexplored coast. 

During the autumn Miss Alice C. Fletcher found it nec- 
essary to revisit Oklahoma for the purpose of completing 
the ritual of the Pawnee ceremony known as the Hako, 
of which the greater portion was collected during the last 
fiscal year. In connection with the collection of this 
material she was fortunate in obtaining also much addi- 
tional information touching the ceremonial and ritualistic 
life of this highly interesting and little -studied tribe. 



Work in Esthetology 

In addition to administrative duties in the office as 
Ethnologist in Charge and part of the time as Acting 
Director, and the field work noted, Mr McGee engaged 
in researches relating to the primitive symbolism found 
among the American aborigines and other lowly peoples. 
Certain symbols are of nearly world-wide distribution, 
and extend into several stages of culture — for example, 
the swastika, or filfot, appears on all of the continents 
except Australia, and its culture range extends at least 
from higher savagery into the lower strata of civilization. 
Before the extremely wide range of such symbols was 
ascertained various inquirers were led to regard their use 
as an evidence of cultural identity, and hence of the orig- 
inal unity of the peoples among whom they were found ; 
but since they have been observed among highly diverse 
peoples in different stages of culture and on remote con- 
tinents this interpretation has been modified or aban- 
doned in large measure, and students have set themselves 
to the task of tracing the development of the symbols in 
particular cases. The recent researches have shown that 
symbols of quatern character, like the swastika, express 
or reflect modes of thought especially characteristic of 
lower (but not lowest) culture, yet extend well into civ- 
ilization and enlightenment. At the same time the 
researches bring to light such diversities in the nature 
and applications of the concepts expressed by the symbols 
as to indicate, if not demonstrate, independent develop- 
ment. Thus, quatern symbols abound among the Papago 
Indians of Arizona and Sonora, as well as among several 
neighboring tribes, yet the Papago concej)t is distinct, as 
is shown by its extension to time as well as space, this 
extension carrying such archaic features of ritual and 
ceremony as to indicate increasing independence of the 
concept in the generations traced backward. 


The neighboring Zniii Indians have a more highly 
differentiated concept in that their u Cult of the Quar- 
ters ' involves six directions (zenith and nadir in addi- 
tion to the cardinal points) , yet the symbol retains the 
original quatern form, with two added elements so placed 
as to destroy the symmetry of the figure. These in- 
stances of diversity in symbol, and still greater diversity 
in meaning of the symbol (or in the primary concept) , 
might be multiplied almost indefinitely ; they merely give 
some indication of the development of simple quatern 
symbols and of the complex and protean magma of 
thought out of which they have been developed by simple 
processes and easy steps. Incidentally the examples 
marshaled corroborate and extend the law of activital 
coincidences formulated in an early report of the Bureau; 
but the applications of the recent study are numerous and 
useful, especially in their bearing on symbolism in gen- 
eral and on the development of systems of counting. The 
results of the study are incorporated in the Nineteenth 
Report in the form of a brief paper entitled ' 4 Primitive 

During the earlier portion of the year Dr Fewkes ar- 
ranged for publication a series of graphic representations 
of the personages composing the Hopi pantheon, together 
with full descriptions of the pictures and a discussion of 
characteristic paraphernalia of the personages repre- 
sented. The representations are in outline and color and 
well illustrate the early stage in the development of 
graphic art reached by the more advanced among the 
aboriginal tribes ; hence they throw strong light on the 
codices and other pictorial essays of the more southerly 
tribes, especially those of Mexico, Central America, and 
Peru. The pictures were executed by a native artist, 
who was also a priest in the hieratic or sacred organiza- 
tion through which the tribal mythology is maintained, 
and each picture is a faithful reproduction of ancient 
representations handed down through many generations. 
The material has been published in the Twenty -first An- 
nual Report, the original drawings being used as copy and 


reproduced in slightly reduced facsimile. The work is 
deemed an important contribution to knowledge of the 
aborigines in several respects. It illustrates the motives 
and conventions of aboriginal art in both form and color ; 
it reveals the role of symbolism in primitive art with re- 
markable clearness; it illustrates with satisfactory com- 
pleteness the nature and structure of a typical barbaric 
pantheon; and since the symbols and conventions (and, 
indeed, the personages represented) are of great constancy 
in primitive thought, it affords a series of types available 
for use in identification and comparison of a wide range 
of symbolic representations among the Pueblo and other 
tribes, not only in ceremonies and sacred paraphernalia, 
but in the decoration of fictile ware, basketry, woven 
fabrics, etc. 

Later in the year Dr Fewkes was occupied with a sys- 
tematic study of the collections made by him in Arizona 
and New Mexico during 1896 and 1897, the study being 
carried forward with special reference to the symbolic 
decoration of the fictile ware. All systematic investi- 
gators of the decorative devices used by primitive peoples 
have been impressed by their constancy, that is, by 
the exceeding slowness of modification. They have also 
been impressed with the dependence of the modification 
on external forces and conditions rather than on the 
spontaneous internal factor so prominent in the art of 
advanced culture. Recognizing these characteristics of 
primitive art, Dr Fewkes undertook to define the symbolic 
(or esthetic) types prevailing among the peoples of Walpi, 
much as a naturalist might define types of animal and 
vegetal life for the establishment of species and genera 
and orders, and for tracing the lines of vital development 
in a distinctive environment. His symbolic types were 
based on specimens observed among the tribesmen or 
obtained from sites occupied by their ancestors during the 
historical period; and he soon found that the types served 
to indicate what may be termed a symbolic province, that 
is, a region throughout which the symbolic devices were 


similar, but in which they differed essentially from those 
of other regions. In this way he defined an ethnic dis- 
trict and established standards for the guidance of future 
investigation and also for the localization of ill -labeled 
specimens in museums; for many collectors have been 
content to label specimens of symbolic pottery and other 
objects " Arizona,' " Pueblo region,' or by other large 
and indefinite political or natural divisions, thereby 
confusing important symbolic distinctions and ethnic 

As his investigations of the symbolic types progressed, 
Dr Fewkes became more deeply impressed than any pre- 
decessor with the persistence of motives and the regular- 
ity of their evolutional lines; and he conceived, in a 
definite and constructive way, the possibility of tracing 
prehistoric migrations by means of the decorative symbols, 
that is, of employing symbolic devices as prehistoric rec- 
ords, reading from them the tale of tribal movements 
before the coming of Coronado — he conceived the possi- 
bility of coordinating the archeologic record as taught by 
symbols with tribal traditions, and the double advantage 
of mutual verification between tradition and symbolic 
record. Proceeding in accordance with these ideas, he 
obtained from living Hopi men traditions of a former 
residence of their ancestors at a locality which they called 
Homolobi, and by excavations he identified this site and 
verified the traditions, thereby extending his knowledge 
of the evolution of the symbolic types ; for the Homolobi 
collections (now in the National Museum) not only abound 
in decorated ware, but are notably rich in symbols suscep- 
tible of interpretation. Subsequent exploration brought 
him to the site of a ruin on Chevlon creek, where excava- 
tion revealed another stage in the same general line of 
symbolic development, which corroborated the vague and 
shadowy tradition that Hopi clans once inhabited this site. 
He later sought a locality noted in still vaguer migration 
legends, and was gratified by finding near Chaves pass 
the archeologic record of this stage in migration inscribed 


in symbols related to the higher type from the more 
northerly localities. 

Beyond this point ruins which mark traditional halting 
places in migration were not located ; beyond it the sym- 
bolic development has not yet been traced ; but there is 
good ground for anticipating that when Dr Fewkes returns 
to the field he will obtain still earlier records of the pre- 
historic movements and development of this branch of 
the Pueblo peoples. The work is deemed of much impor- 
tance as a verification of aboriginal tradition, as a means 
of verifying other migration legends, and as a promising 
introduction to the practical interpretation of history 
unwittingly recorded in graphic symbols. Incidentally, 
the work corroborates the earlier conclusion reached in 
the Bureau, that the Pueblo peoples are a resultant prod- 
uct of Southern culture and Northern blood; yet the sig- 
nificant details throw new light on the entire problem. 
The report is elaborately illustrated by colored photo - 
grajjhs of the ware from the several localities examined ; 
it occupies a portion of the present volume. 

Work in Technology 

The earlier accounts of exploration in the territory 
occupied by the Cocopa Indians seemed to indicate that 
the tribesmen occupied the coast of Gulf of California 
and were of maritime habits; but in the course of the 
expedition led by the Ethnologist in Charge it was defi- 
nitely ascertained that the folk are essentially agricultural 
and confined, at least so far as habitations are concerned, 
to the interior. The industrial condition of the tribe was 
found to be of much interest. The tribal habitat com- 
prises the Lower Colorado valley from the International 
Boundary southward to the head of the gulf, together 
with a few tributary valleys descending from the Cocopa 
mountains on the west. The main valley is broad and 
diversified by distributaries, or bayous, of which the most 
important is Hardy river, or "Hardy's Colorado." There 
are also several fairly permanent basins, filled by the 


annual floods and slowly evaporated during succeeding 
months, and the greater part of the broad bottom is 
swept by. the freshets. Within the region lie a number 
of "mud volcanoes," apparently analogous to the "mud 
lumps ' of the Lower Mississippi, which have attracted 
much attention by reason of their novelty, though they 
are quite subordinate to the general features. The entire 
district affords the closest American parallel to the valley 
of the Nile, not only in physical conditions, but in the 
influence of these on human conditions. 

Like northern Africa, the general region is one of 
extreme aridity, the rainfall (averaging less than 2 inches 
yearly during the last quarter -century at the typical sta- 
tion of Mammoth Tanks) being negligible; while the 
habitable district is well Avatered by annual freshets of 
remarkable regularity in period and height. These fresh- 
ets not only flood but fertilize the riparian lowlands ; they 
control directly the local flora and somewhat less directly 
the local fauna, and they regulate the movements, most 
of the industrial habits, many of the social customs, and 
much of the mythology of the human population. Dur- 
ing the greater part of the year water is obtainable only 
from the shrunken river, on whose banks grow most of 
the seed-bearing and root-yielding plants available as 
food, so that the people are led to occupy the lower bot- 
tom lands. Here the cultivated crop plants are sown in 
soil soaked by the flood and enriched by its silt deposit, 
to grow and ripen rapidly under the subtropical sun; 
here habitations are erected, naturally of light and tem- 
porary character, and here the small and scattered vil- 
lages characteristic of the tribe grow up during each late 
summer and early autumn. The chief crop plants are 
corn (maize), beans, peas, squashes, and melons, and it is 
noteworthy that most of these represent the aboriginal 
plant stocks brought under cultivation in pre-Columbian 

Fishing and hunting the abundant waterfowl, as well as 
other game, contribute to the tribal subsistence, and dur- 
ing recent years part of the crop of corn, beans, and peas 


is carried on horseback to Yuma and bartered chiefly for 
appareling. Early winter is the time for ceremony with 
the attendant feasting, and by early spring, when the 
greater and less portable part of the annual crop is con- 
sumed, the families prepare for the annual migration to 
the higher lands, where they await the rise and subsi- 
dence of the vernal flood. On its passing they return to 
the low grounds, to rebuild and plant on the last year's 
farms or elsewhere according to the changes wrought by 
the freshet or the chance of death and mortuary observance. 
Naturally an agriculture depending so largely on chance 
conditions is improvident, comparatively unproductive, 
and incapable of sustaining any considerable or concen- 
trated population, so that its tendency combines with 
that of annual migrations to stifle the home sense and to 
scatter the members of consanguineal groups, and thus 
to affect the social organization. The recurrent floods also 
affect the ceremonies and attendant faiths of the tribes- 
men in various ways; for example, they control mortuary 
observances and have undoubtedly led indirectly to the 
custom of burning the bodies of decedents in and with 
their houses, distributing their property to nonrelatives, 
and incidentally destroying adjacent houses and other 
property. This dispersive social factor combines with 
that growing directly out of the agricultural methods, and 
not only prevents the development of village life with the 
concomitant institutions, but perpetually impoverishes 
the tribe. Thus the Cocopa Indians present an industrial 
paradox, for while they occupy one of the garden spots of 
the Western Hemisphere, whose natural freshets might 
be so utilized as to sustain an enormous population, they 
subordinate themselves to the environmental conditions 
and remain one of the poorest and most hopeless of the 
American tribes. 

During the earlier part of the year Dr Albert E. Jenks 
(then a correspondent of the Bureau) revised his memoir 
on u The Wild Rice Gatherers of the Upper Lakes' (pub- 
lished as part of the Nineteenth Annual Report), incor- 
porating some of the results of recent researches. On 


June 1 lie was appointed to the position of Assistant 
Ethnologist in the Bureau, and was assigned to work 
related to his previous researches. He at once took up 
the subject of birch bark, with the aboriginal industries 
depending on this natural commodity of a considerable 
fraction of the North American continent. One of the 
most important products of the birch -bark industry is 
the canoe ; and this, like other industrial products of con- 
sequence, exerted a powerful influence on the lives of the 
producers. Through one of those harmonies of nature 
on which the progress of mankind so largely depends, 
much of the birch -bearing region of North America (a 
zone stretching from Maine to Washington State and 
Alaska, and extending from below the Great Lakes nearly 
to the treeless Arctic) is also the region of late, Pleisto- 
cene glaciation, and hence of glacial lakes, swamps, and 
labyrinthine streams; so that throughout the period of 
aboriginal development an ideal canoe material coexisted 
with illimitable functions for the canoe in the way of 
travel and transportation. 

Under the natural combination, aided by native intelli- 
gence and skill, the lakes and streams became routes of 
passage, and by reason of the lightness and strength of 
the material, and the lowness and narrowness of the ice- 
molded divides, portages were easy, so that the routes 
passed from lake to lake, river to river, and drainage sys- 
tem to drainage system, practically across the continent. 
Under the stimulus of facility, the birch -canoe makers 
became travelers and explorers; energetic hunters and 
fishermen explored new waters and carried tribal knowl- 
edge into new regions ; ambitious scions struck out into 
the remoter wilderness to make conquest over the unknown 
and often to establish families and clans, and eventually 
tribes, in new localities ; so that in course of time the pad- 
dlers of the light canoe carried their kindred, their dia- 
lects, their faiths, over the greater part of the vast region 
defined by the birch tree and the glacial waterways. Most 
of the canoemen belong to the Algonquian stock, most 
of the remainder to the Athapascan stock ; and the recent 


researches render it clear that their water craft was a 
leading factor in determining their wide distribution and 
their success in making conquest of the continent up to 
the plane of aboriginal standards. The detailed results of 
the work are in preparation for an early report. 

In tracing the joint lines of migration and esthetic 
development noted in other paragraphs Dr Fewkes be- 
came impressed with the fact that among the ancestors 
of the Hopi Indians the esthetic standards were much 
more permanent than the industrial standards. Through- 
out the entire course of tribal migrations retraced by his 
researches — a course covering several distinct treks, alter- 
nating with periods of stable settlement, the whole cov- 
ering some centuries — the symbolic devices inscribed on 
the fictile ware remained constant or underwent only 
slight and easily traceable modifications, while at each 
successive settlement new materials were utilized in the 
pottery making, the manufacturing processes and the 
final forms of the ware being manifestly adjusted to the 
character of the material. The discovery that the indus- 
trial activities (which directly measure the conjustment 
of man and environment) are the most progressive of the 
entire series is not, of course, novel; still less is it novel 
to learn that the especially conservative esthetic concepts, 
which are at once hereditary and prophetic, as shown by 
Groos, outlive whole generations of contemporaneous 
industrial concepts ; yet the example is notably apposite 
and instructive, largely by reason of the freedom of the 
folk from external interference, with the consequent sim- 
plicity and integrity of the record. The details are incor- 
porated in Dr Fewkes's paper in another part of this 

In the course of his reconnoissance of central and 
-outhern Arizona Dr Frank Russell gave especial atten- 
tion to the architectural features of the ruins, and defined 
a number of types, of which one or two are new to south- 
western archeology. The work was still in progress at 
the close of the fiscal year. 


Work in Sociology 

A portion of the year was employed by the Director 
and the Ethnologist in Charge in reviewing the abundant 
data in the Bureau archives relating to aboriginal insti- 
tutions, and in systemizing the principles of sociology in 
the light of these data. One of the lines of inquiry, ren- 
dered important not only by inherent interest but by 
current problems growing out of the recent expansion 
of the territory of the United States, relates to slavery 
among the primitive peoples, and the researches render 
it clear that the relationships so designated vary widely 
with intellectual plane or culture grade — indeed, the 
social subordination of lower culture is so unlike the 
slavery of civilization that the application of the same 
designation to both institutions is quite misleading. In 
the slavery of civilization the slaves are not only aliens 
but chattels, personal ownership of whom is definitely 
established and maintained through laws relating to 
tenure, bequest, conveyance, etc., but in savage society, 
in which personal proprietary rights are inchoate or non- 
existent, in which the tenure inheres practically or abso- 
lutely in the group, in which bequest is hardly, if at all, 
recog;nized, and in which thrift sense is lacking and prop- 
erty sense involved with mythic factors, such slavery is 
simply impossible. True, there are many recorded in- 
stances of slavery among lower tribes, but most of these 
rest on casual or superficial observation, or on other testi- 
mony stopping short of inquiry into the precise nature of 
the relations between the supposed slaveholders and the 
supposed slaves, while the convenience of the common 
term for the expression of social inequality has contrib- 
uted to mislead recorders and (still more seriously) 

To understand the so-called slavery of savagery it is 
necessary to grasp the mode of social organization char- 
acteristic of that culture grade. As shown chiefly through 
the researches among the American aborigines, such 
organization is based primarily on consanguinity (actual 


or imputed) , and secondarily on age ; and the relations 
growing out of these factors are kept constantly in the 
mind of every member of each clan and tribe by habitual 
forms of address. So the constituent individuals of a 
given clan are fathers and mothers, sons and daughters, 
brothers and sisters, and these relationships are constantly 
indicated in salutations, and even in ordinary conversa- 
tion (the precise relationship to the speaker" being com - 
monly expressed also by a pronominal element) . At the 
same time it is constantly borne in mind that father and 
son, mother and daughter, are not coordinate, the for- 
mer being the superior by reason of greater age ; similarly 
brethren are classed as elder brothers and younger broth - 
ers, while the female kindred of the same generation are 
classed as elder sisters and younger sisters, and the elder 
are always deemed superior, the younger inferior, in rank. 
By simple and practical extension of the system, the rela- 
tive ages of all persons in the clan are kept in mind ; and 
since, according to the universal usage of savagery (so far 
as known), superior age confers authority, there is a prac- 
tically simple, though theoretically complex, regimenta- 
tion running through the entire clan, whereby the eldest 
person commands all and obeys none, while the youngest 
person obeys all and commands none, and each other per- 
son is entitled to command and bound to obey in the direct 
proportion of relative age. This regimentation is com- 
plicated by various factors, such as adoption, and (espe- 
cially) what may be called promotion and demotion, that 
is, advancement in u age" (rank) by common consent in 
recognition of prowess, etc., with correlative reduction in 
u age ' ' as the penalty for cowardice, etc., so that the actual 
age relations may be completely lost; yet the imputed 
relationship serves practical purposes, and the organiza- 
tion is maintained with unimpaired efficiency by means of 
relationship terms. The same system is extended from 
the clan to the tribe, in which the several clans are 
ranked in the order of " age " (of course imputed), and 
eventually to the tribes united in confederacies; so at last 


the system reaches every member of the tribal confed- 
eracy and each is entitled to command or bound to obey 
any other according to the relationship expressed in the 
form of salutation and constantly kept alive in conversa- 
tion. True, uncertainties and differences of opinion may 
arise, especially between the remoter individuals and 
groups ; commonly these are settled by more or less pro - 
longed deliberation and discussion, or "council,' though 
some of the bloodiest wars of Indian history grew out of 
such misunderstandings; yet even the appeal to force 
and arms but serves as a means of settlement of the dis- 
pute, for the conquerors thereby become the elder and 
the conquered the younger in primitive thought. So, 
too, when stranger tribes meet, both are constrained by 
universal tribal law, and proceed to council or war, as the 
case may be, for the purpose of fixing the relative " age," 
with the consequent right of command, and in some cases 
the question may remain open for centuries (as between 
the Apache and the Papago) and lead to interminable 
warfare. Now, the conquered tribe may merely retire 
from the field of dispute, leaving what' both conceive to 
be the verdict of superhuman potencies beyond reach" of 
continuous execution ; but if the contestants are actually 
related, or if the conquest is complete, they commonly 
remain in association, the survivors of the conquered 
families being absorbed or more formally adopted into 
the conquering tribe, and perhaps distributed among the 
families of that tribe, whereupon all the captives become 
subordinate to each and all of the conquerors, to whom 
thenceforth they owe obedience. Commonly it is this 
condition of obedience on the part of a certain class or 
group to the commands of another class or group which 
impresses observers and leads to the records of slavery 
among primitive folk, though the institution involves no 
ownership of human chattels, no rights or duties save 
those connected with a system of rank correlated with 
relative age, actual or imputed. The institution might 
better be styled wholesale adoption, or collective adop- 
tion, than slavery. Among the American aborigines the 


captives, or adoptees, are usually assigned an u age" cor- 
responding with the time of their entry into the tribe, 
so that they are compelled thereafter to obey all children 
then living, and are entitled to command all children sub- 
sequently born into the tribe, and there is thus a fixed 
way whereby they attain in time the rank of the con- 
querors. Moreover, the method of promotion permits 
any "slave' (that is, captive -junior) to attain "age" by 
the display of prowess, industry, skill, generosity, or 
other attributes appealing to the sentiments of primitive 
men. Among certain other peoples, the custom of collec- 
tive adoption appears to be so modified that the captives 
remain juniors not only to members of the captor tribe 
born anterior to the captivity, but to all others, and it is 
this modified institution which matures in actual slavery 
with the development of property -sense ; but even in this 
case there are (at least in the early stages) devices for the 
manumission or liberation of, or the acquisition of rank 
by, captives (or captive -descendants) of exceptional abil- 
ities. The several primitive customs grade into the insti- 
tution of slavery proper in ways which are of much 
interest, but which need not now be followed ; it suffices 
to emphasize the important distinction between the captive 
subordination of primitive peoples and the real slavery of 
some civilized nations. 

In the course of his researches among the Cocopa In- 
dians Mr McGee discovered several industrial factors of 
dispersive tendency, that is, factors tending to weaken 
home ties and family bonds and to scatter the families 
and clans; and naturally these factors are reflected in 
the social organization. The tribe is now distributed 
over an area of several thousand square miles, extending 
from the International Boundary on the north to the head 
of salt water (of Gulf of California) on the south, and 
from the eastern border of the Colorado bottom to the 
base of Cocopa mountains ; and within this area are seven 
subtribes, of which some, and perhaps all, are really 
clans, each organized under a subchief and all definitely 
united under a head chief, the present incumbent of this 


office being a man of parts, an orator of ability, and a 
leader of much shrewdness, commonly known as Pablo 
Colorado. Now, naturally (and necessarily for the main- 
tenance of tribal integrity) the dispersive factors are 
counteracted and balanced by connective factors; and 
while it is probable that some of these remain undiscov- 
ered, a few of no small significance were detected by Mr 
McGee. As has been mentioned, the mortuary observ- 
ances include sacrifice of all the immediate belongings of 
decedents, for immediately after the death of a tribes- 
man his personal possessions — horse, saddle, weapons, 
implements, apparel, grain and other food stuffs, bed- 
ding, dogs, etc. — become public and are distributed 
among nonrelatives in the order of arrival, while any 
unclaimed residue is burned with the body and house. 
Several social consequences attend this industrially 
improvident procedure. In the first place, the largess is 
an incentive to maintaining connection between the scat- 
tered families and clans and to lively (albeit morbid) 
interest in the state of health of invalids, thrifty pro- 
ducers, and other members of the tribe; again, the actual 
mortuary distribution brings together scattered tribesmen 
and their families and unites their interests in ceremonies 
of affecting if not imposing character; and finally the 
material sacrifice commonly leaves dependents (widows, 
children, and perhaps agelings) to be supported by the 
informal public bounty of tribal life, or perhaps to be 
distributed among scattered families in such manner as 
to strengthen sentiments of communality and to keep 
alive the sense of community in interests. This factor is 
prominent in the customs of the tribe, and its influence is 
direct and easily traceable. 

A less direct factor of similar tendency is found in the 
marital customs, or rather in the observances preceding 
and preparing the way for marriage. The girls' puberty 
feast is, indeed, one of the most imposing and widely 
heralded of the tribal ceremonies; commonly it brings 
together representatives of all the subtribes or clans ; and 
the proceedings are conducted with extreme formality 


and dramatic impressiveness. The principal ceremony 
lasts through a night, following a day of preparation and 
followed by another day of final feasting, accompanied 
by games, etc. The central episode is the temporary 
burial of the novitiate; a shallow pit is excavated, and in 
this a fire is made, as for a fish bake; after the earth is 
thoroughly warmed the remaining fuel and coals are re- 
moved, the girl is placed in the pit and buried to the neck 
with the earth thrown out in making the excavation; 
there she spends the night, and in the morning is extri- 
cated and brought before the assembled tribesmen as a 
woman; and commonly a match is made with a repre- 
sentative of some more or less remote branch of the 
tribe. Through the ceremony community of thought is 
maintained in most effective fashion, and through the 
resulting marriage the two families are united to the 
extent that a common consequence is the breaking of 
a new path, often many miles in length, through the 
luxurious herbage of the annually flooded bottom land. 
The formal organization of the Cocopa tribe is in large 
measure esoteric, so that it can be ascertained fully 
only after prolonged and intimate acquaintance with the 
tribesmen, but the preliminary investigation serves to 
show that the field of inquiry is one of promise. 

In his comparative study of myths Mr J. N. B. Hewitt 
found various references to social customs of such sort as 
to indicate clearly certain archaic institutions of the Iro- 
quoian Indians. Thus the Onondaga legends illumine 
the legislative and executive customs of the tribe, and, 
while ostensibly giving traditional warrant for the cus- 
toms, they really picture a somewhat earlier stage in the 
development of institutions than that found by the Cau- 
casian pioneers. In this tribe all matters of public policy, 
especially the selection of chiefs and the discontinuance 
of war, were first considered by the elderwomen in fairly 
definite clan councils. Their conclusions were formally 
communicated to a male spokesman, usually the elder 
brother (actual or putative) of the elderwoman, and by 

22 ETH — 04 III 


this spokesman, with others of similar character from 
the other clans, the opinions of the mothers were brought 
before the exclusively masculine tribal council for debate 
and final decision. In this way the women sitting in clan 
council constituted the primary legislative body, while 
their brothers sitting in tribal council formed a senate or 
final legislative body whose decisions were binding on the 
executives of clans and tribes; so that the social organi- 
zation may be classed as adelphiarchal (like that of the 
Seri Indians described in earlier reports) in principle, 
though largely patriarchal in detail. As among the Seri, 
too, the maternal features of the legislation were paral- 
leled by recognition of large maternal rights in material 
possessions — for example, throughout the Iroquoian tribes 
the control or nominal ownership of lands was in the 
women as the collective and perpetual mothers of the 
tribe. These and other points of general interest are set 
forth in Mr Hewitt's memoir, which was assigned to the 
Twenty -first Annual Report. 

Work in Philology 

Throughout a considerable part of the year the Director 
was occupied in developing and applying the system of 
linguistic classification foreshadowed in the last report. 
Primarily, languages- are devices for the expression of 
thought; secondarily, they are mechanisms for shaping 
thought. The simplest languages are emotional and 
largely demonstrative, comprising not only articulate vocal 
utterances, but inarticulate sounds, gestures, facial ex- 
pressions, etc., and these spontaneous expressions of feel- 
ing and thought grow into the four leading lines of lin- 
guistic development. The simplest of these is gesture 
language (or sign language) , which arises largely in pan- 
tomime, but matures under favorable conditions in highly 
complex systems such as those investigated by the late 
Colonel Mallery and more recently by Major H. L. Scott 
(whose studies were unfortunately interrupted by the 
Spanish -American war) . A far more important line of 


linguistic development is that of oral speech, and the ac- 
tivities of expression have been so long and so vigorously 
exercised in this line as to have developed a series of spe- 
cial organs differing widely in refinement of function and 
delicacy of structure from those of lower animals. By 
means of these organs the speaking animal, Man, gains 
mastery of sound, which is created at will and reduced to 
vocables, tones, and sentences in such manner as to convey 
ideas of the utmost complexity with hardly perceptible 
loss of meaning ; and with the development of words and 
sentences lexicology and grammar arise, while etymology 
and sematology gradually acquire importance. The third 
line of linguistic development is that of written language, 
which first involved manual adaptation, together with a 
revolution in mode of thought, and afterward involved the 
invention of that long series of mechanical devices now 
forming the sign and measure of higher intellectuality. 
The last line of linguistic development is that represented 
by characters expressing quantitative values ; it may be 
styled logistic language . Although based primarily on the 
rich records of. aboriginal American languages preserved 
in the archives of the Bureau, the system of linguistic 
classification was shaped by extended comparisons with 
the various languages of Europe and Asia, together with 
some of those of Australia, Africa, and Polynesia. The 
system was freely discussed with students and published 
in preliminary form for the purpose of eliciting further 
suggestion and criticism ; the matter was incorporated in 
full in the Twentieth Annual Report. 

In connection with the linguistic classification, the 
Director continued to study the recorded languages of the 
Mexican and Central American tribes, with a view to 
classifying these tribes by linguistic affinities in a manner 
corresponding to that already adopted for the American 
tribes north of Mexico (the classification being published 
in the Seventh Annual Report). In this works he had 
the constant assistance of DrCynis Thomas, whose famil- 
iarity with the literature of the southern districts of North 


America proved invaluable. Before the end of the year 
a preliminary classification was made and mapped ; but it 
is deemed unwise to submit the matter for publication 
pending reexamination of various critical points. It has 
been the good fortune of the Bureau to see its classifica- 
tion and mapping of the tribes north of Mexico adopted 
widely, and it is naturally desired that the continua- 
tion of the work southward shall be equally worthy of 

Dr Albert S. Gatschet continued the arrangement of 
the comparative Algonquian vocabulary, and also carried 
forward his analysis of the complex structure of the 
Peoria language. In both directions his progress was 
considerable and his results of much value, not only as 
an aid in formulating the linguistic classification above 
described, but to the collaborators of the Bureau and 
students generally. 

Dr Franz Boas continued the arrangement of linguistic 
material for publication at intervals throughout the year. 
In addition, he revised the proofs of his memoir entitled 
" Kathlamet Texts,' submitted just before the close of 
the last fiscal year and transmitted for publication in bul- 
letin form early in the present year. By reason of the 
highly technical character of the matter, composition was 
necessarily slow and proof reading laborious; but the 
matter is now all in type. 

The Natick Dictionary, compiled from the Eliot Indian 
Bible by the late James Hammond Trumbull (noted in 
the last report) , is still in the printer's hands, though 
nearly ready for publication. 

In connection with the collection of Iroquoian myths, 
Mr Hewitt has continued recording the vocables and 
working out the grammatic structure of the languages 
spoken by several Iroquoian tribes. Some of the results 
of the work appear in his memoir on comparative myth- 
ology in the Twenty -first Annual Report of the Bureau; 
others are in condition for incorporation in future reports. 

As already noted, Mr John R. S wanton spent the entire 
year in collecting linguistic material in British Columbia. 


The languages of this district give promise of special 
importance in their bearing on questions of tribal migra- 
tions and intertribal relations. Mr S wanton has not yet 
taken up the preparation of his material for publication. 
The work on the Diccionario de Motul, described in the 
last report, is still under way. A considerable portion of 
the manuscript in Maya and Spanish was transcribed by 
Miss Jessie E. Thomas during the year, and Senor 
Auclomaro Molina, of Merida, Yucatan, is engaged in fur- 
nishing an English translation and in extending the 
vocabulary through personal acquaintance with the Maya 

Work in Sophiology 

As has been indicated by the contents of previous 
reports, the Director has for some years been engaged 
in developing a system of anthropologic classification 
designed primarily to serve as a basis for the researches 
in the Bureau, though it is hoped that the system will be 
of use to the students of the Science of Man throughout 
the world. It was the partial development of this system 
that led first to discrimination of the human activities 
and later to the definition of the five groups of activities 
observed in the researches and described in recent reports. 
During the last five years several of the groups or cate- 
gories of activities have been formulated and character- 
ized with some degree of fullness. The treatment began 
with the arts, or esthetic activities, and proceeded to the 
industries, or technical activities, and thence to the insti- 
tutions expressing social activities. During the past year 
the characterization was extended to languages, or the 
activities designed for expression, as already set forth, 
and toward the end of the year the last and most complex 
of the activital groups, that is, the sophic activities 
involved in opinion, together with myth, faith, and the 
more refined and ennobling products of mentation, was 
taken up. Fair x> r °g ress was made in the analytical 


During his Southwestern expedition Mr McGee found 
opportunity to witness certain ceremonies of the Yaki 
Indians, which were of interest partly because the tribe 
has been little studied, partly by reason of the prominence 
of zoic motives in the vocalization and instrumentation, 
as well as in the gestures and movements of the cere- 
monial dance. In portions of the ceremony each actor 
impersonated an animal. He wore a headdress (not 
extended into a mask, as among more northerly tribes) 
consisting of a scalp, with ears, horns, and other append- 
ages of the animal kind, and leggings abundantly deco- 
rated with claws or hoofs of the same animal. He carried 
a rattle or flute, used to imitate the voice of the tutelary 
or the sound of its movements, while he imitated its notes 
of alarm, fright, pain, and pleasure with his own voice, 
and mimicked its corresponding movements ; yet in other 
parts of the ceremony the same actors passed by carefully 
graded stages into the strictly conventional movements of 
a dance involving collective action of considerable com- 
plexity. Briefly, the ceremony seemed to be character- 
ized by a remarkable combination of symbolic and con- 
ventional features, indicating an exceptional range from 
the primitive impersonation to the formal figures and 
movements attending moderately advanced culture. 

Mr James Mooney continued his researches relating to 
the mythology of the Cherokee Indians, making good 
progress in the collection of additional material in the 
field, as well as in the extension of comparisons between 
the myths of the Cherokee and those of other tribes and 
peoples. The application of comparative study to primi- 
tive mythology is proving highly instructive and useful. 
In the infancy of ethnologic research students were fre- 
quently struck by the discovery of activital parallels, or 
similarities, among more or less remote peoples, and were 
led thereby to infer previous contact, or even closer rela- 
tionship, between the peoples; but as study progressed 
and new parallels were discovered, even among the 
remotest peoples of the earth, the verity of the inference 
came to be questioned, and finally the law of activital 


coincidences was formulated as a convenient generaliza- 
tion of the facts connected with independent development 
of devices produced in the constant adjustment of the 
intelligent organism to its environment. At first the law 
of activital coincidences rested chiefly on industrial arti- 
facts ; then it was found to have equal support in the 
esthetic products of various peoples ; next it was found 
to have still stronger and more direct support in institu- 
tions, in the devices and features of social organization; 
while certain features of language were found also to 
indicate the extent and efficiency of coincidental interac- 
tion between mind and nature in shaping the activital 
products. Hitherto most investigators of mythology 
have .been content with discrete studies and explorations, 
or, at most, with exoteric parallels. Accordingly many of 
them have stopped with the inference of former contact 
or kinship on which the students of industrial artifacts 
rested a quarter century ago, that is, their studies were such 
as to bring out resemblances among the mythic systems 
examined, but not such as to detect and properly empha- 
size the essential differences. Now, Mr Mooney's com- 
parisons, although not exhaustive, are sufficiently general 
to permit discrimination of the exoteric coincidences from 
esoteric motives in the myths. Accordingly they clear 
the way for the application of the law of activital coinci- 
dences to primitive mythology, if not to sophiology in 
general. The greater part of the material completed for 
publication has been incorporated in the memoir on 
"Myths of the Cherokee," published in the Nineteenth 

Another comparative study of myths has been carried 
forward by Mr J. N. B. Hewitt; and this investigation 
is noteworthy in that the comparisons are confined to a 
limited group of confederated tribes (of the Iroquoian 
stock) and in that the features compared are in excep- 
tional degree esoteric. The myths were obtained at first 
hand and carefully recorded and verified in the aboriginal 
terminology, after which literal and free translations were 
made, so that each chapter of the work is at once a 


linguistic record and the best obtainable version of the 
ancient traditions. Now, it is noteworthy that most of 
the similarities found thus among the several Iroquoian 
myths are rather external than internal, rather superficial 
than essential, and, concordantly, that the more im- 
portant differences are primarily internal, that is, more 
directly connected with concept and motive than with 
ritual and emblem. The voluminous material was prac- 
tically ready for the press at the close of the fiscal year 
and was assigned to the Twenty -first Annual Report. 

During the closing months of the year Dr Fewkes was 
employed in summarizing his own observations and those 
of others in the pueblo region, with the object of present- 
ing an outline of Pueblo mythology. As noted in earlier 
reports, the pueblo region is arid, and hence infertile and 
harsh as an environment for human inhabitants 1 , and the 
harshness of environments is curiously reflected in highly 
differentiated beliefs and ceremonies, so that the pueblo 
region as a whole may, perhaps, be regarded as a sophic 
province, that is, a province defined by a distinctively 
typical series of myths and faiths. Good progress was 
made in the work, which was not, however, completed at 
the close of the fiscal year. 

In addition to the inquiries connected with the classi- 
fication of the languages of Mexico and Central America, 
Dr Cyrus Thomas gave continued attention to the hiero- 
glyphic records of the inscriptions and sculptures of 
Yucatan and interior Mexico, materially supplementing 
and extending his paper on calendric systems published 
as a part of the Nineteenth Annual Report. He made 
some progress also in the preparation of a final memoir 
on the codices. 

Although seriously handicapped by ill health, Mrs 
Matilda Coxe Stevenson continued the preparation of her 
memoir on the ceremonies and myths of the Zuiii Indians. 
A portion of the manuscript was submitted for editorial 
revision in May, and the remaining chapters were reported 
as nearing completion at the end of the fiscal year. 


As noted in the last report, an exceedingly valuable 
acquisition was made through Miss Alice C. Fletcher in 
the form of the Pawnee ritual known as the Hako ; but 
on arranging the material for printing certain breaks 
were found which seemed of such importance as to war- 
rant postponement of publication pending further efforts 
in the field to complete the ritual. Accordingly Miss 
Fletcher revisited Oklahoma, and afterward brought her 
principal informant to Washington, where the record was 
finally made perfect. The ritual is remarkable for ex- 
tent and fulness, for the clear light which it throws on 
archaic customs and beliefs, and for the systematic and 
harmonious development of the musical and terpsicho- 
rean features. The original record was obtained by aid 
of the graphophone, and this record was then written in 
words and musical notation, and afterward verified by 
repetition. On the whole the ritual is one of the most 
complete ever acquired by the Bureau, and is in every 
way worthy to be regarded as a type of aboriginal ritual- 
istic production. The final arrangement of the material 
was nearly finished at the close of the fiscal year, when 
the work was interrupted by Miss Fletcher's temporary 
absence from the city. 

Work in Descriptive Ethnology 

During the earlier portion of the year Mr F. W. Hodge 
continued the preparation of the Cyclopedia of Native 
Tribes in connection with editorial work, his progress in 
both lines being highly satisfactory. On January 31 he 
resigned his connection with the Bureau to accept a posi- 
tion in the office of the Secretary. The Cyclopedia mate- 
rial was then turned over to Mr Mooney, who has made 
some progress in preparing it for publication . 

During the earlier months of the year Colonel F. F. 
Hilder was, by temporary transfer, engaged in making 
collections in the Philippine Islands under the auspices 
of the Government Board of the Pan-American Exposi- 
tion. After his return he resumed his duties as Ethnologic 


Translator and continued the transcription, translation, 
and annotation of an early Jesuit manuscript history of 
Texas, obtained through the instrumentality of the 
Bureau, but now preserved in the Library of Congress. 
The sketch was found rich in important ethnologic data, 
and the anonymous author was identified by Colonel 
Hilder, through collateral information, as Padre Morn. 
The work was nearly completed when brought to a pre- 
mature end by the sudden death of Colonel Hilder on 
January 21. 


Mr F. W. Hodge continued in charge of the editorial 
work until his resignation took effect, as already noted, 
after which this work was conducted by Mr H. S. Wood. 
The first part of the Seventeenth report and the first part 
of the Eighteenth report were received from the Govern- 
ment Printing Office during the year, and these, with the 
second part of the Seventeenth report, have been distrib- 
uted. The second part of the Eighteenth report was 
not delivered at the end of June, while neither of the two 
bulletins of the new series was quite complete ; and the 
Nineteenth report, though nearly all in type, was not yet 
ready for the bindery at the close of the year. 

Mr De Lancey Gill remained in charge of the illustra- 
tive work, preparing copy for and revising proofs of the 
numerous illustrations for the Eighteenth and Nineteenth 
reports. He also made photo -portraits of some two hun- 
dred Indians, chiefly members of delegations visiting 
Washington in the interest of their tribes, and developed 
a considerable number of negatives made by the several 
collaborators in the field. 


As usual, the several collaborators engaged in field 
operations made more or less extensive collections for 
purposes of study and for ultimate transfer to the U. S. 
National Museum. The largest collection of the sort was 
made by Mr McGee among the Cocopa Indians. It com- 


prised domestic utensils of wood, stone, and clay; sev- 
eral bows with arrows ; war weapons ; complete suits of 
women's apparel; cradles; decorative and symbolic ob- 
jects of shell and bone; flutes, rattles, etc., together with 
the chief vegetal food products used by the tribe, the 
collection being sufficiently complete to permit the con- 
struction of one or more life-size groups. The most 
elaborate war weapon is of interest in that it is designed 
to serve at once as standard and spear, and in that the 
sharpened point for the latter use is at the inner end of the 
shaft, so that the weapon illustrates the centripetal move- 
ment of lowest culture rather than the centrifugal arm 
movement characteristic of advanced culture. Smaller 
collections were made by Mr Mooney among the Cherokee 
Indians, by Mr Hewitt among the Iroquoian Indians of 
Canada, and by Dr Russell in Arizona. A number of 
collections were obtained also by purchase under the more 
immediate direction of the Secretary. Among these may 
be mentioned the Steiner collection of stone implements 
from Georgia, which comprises a large number of types 
and of which a portion was obtained during the last fiscal 
year. Another collection of special note was obtained 
from Major H. N. Rust, of Pasadena, California. It com- 
prises several tyyjes and numerous examples representing 
the stone artifacts of southern California. Advantage 
was taken also of the opportunity to acquire a number of 
the remarkably faithful Indian portraits executed by Mr 
J. H. Sharp, of Cincinnati. A particularly instructive 
collection of obsidian blades (including the largest known 
specimens) was also obtained during the year through Mr 
Nathan Joseph, of San Francisco, while a few particularly 
fine pieces of aboriginal Alaskan workmanship were 
obtained from Lieut. G. T. Emmons. A small collection 
of basketry produced by the renegade Apache at Palomas 
was picked up by Mr McGee, together with several pieces 
of Pima basketry made near Maricopa. A small but 
noteworthy object obtained was an authenticated Sitting 
Bull belt of beaded elk skin; and half a dozen small col- 
lections of stone implements and weapons were secured. 



The property of the Bureau is practically limited to (1) 
office furniture and apparatus, (2) ethnologic manuscripts 
and other original records, (3) photographs and drawings 
of Indian subjects, (4) a working library, (5) collections 
held temporarily by collaborators for use in research, and 
(6) undistributed residue of the editions of the Bureau 
publications. The fiscal year witnessed little change in 
the amount or value of the office property. The accumu- 
lation of manuscripts and other records of original work 
progressed steadily; about a thousand photographic neg- 
atives, together with several hundred prints and a num- 
ber of drawings, were added to the collection of illustrative 
material. The library maintained normal growth chiefly 
through exchange, and the number of back reports was 
considerably reduced through the constantly increasing 
public demand for ethnologic literature. Mr J. Julius 
Lund continued in charge of the property as custodian. 


On January 21, 1901, the Bureau suffered a grievous 
loss in the death of Colonel F. F. Hilder, Ethnologic 

Frank Frederick Hilder, soldier, geographer, and eth- 
nologist, was born in Hastings, England, in 1836. Edu- 
cated at Rugby in the approved manner of the times, he 
afterward graduated from the British military school at 
Sandhurst, and entered the army as a cornet in early 
manhood, at a time when the eyes of all England were 
turned on India. Sent immediately to aid in quelling 
the Sepoy rebellion, he soon saw service of such severity, 
and met it with such intrepidity, that he was awarded 
the Indian Mutiny medal, with special -service bars for 
Delhi and Lucknow. 

It was during this period of his career that Hilder 
traversed the Indo-Gangetic plain, trod the Himalayan 
foothills, and visited the provinces and cities of the 


northwestern empire from Bombay to Kashmir, and 
from the Punjab to Nepal, laying the foundation for a 
broad yet precise geographic and ethnologic education; 
and some of the lectures of even the latest years of his 
life drew inspiration and significant detail from the 
researches enlivening these early campaigns. He saw 
service also in Farther India, Borneo, and the Philip- 
pines, and after rising through a lieutenancy to the rank 
of captain was transferred to Africa. Here he won the 
Egyptian medal, and his skill as military expert and 
organizer attracted such attention that after his return to 
his regiment in India he was recalled and promoted to a 
colonelcy at the express request of the Khedive. 

In Africa, as in India, Colonel Hilder seized every 
opportunity for scientific research ; but his tenure in the 
Egyptian army was cut short by the terrible experience 
of a sand-storm, which so injured his eyesight that he 
decided to abandon a military career. Coming to Amer- 
ica on his recovery, Colonel Hilder met again the conta- 
gion of military spirit stimulated by our civil war, and 
did special work of importance in the Engineer Corps, 
but held so firmly to his election of a peaceful life as to 
decline an American commission. In the later sixties he 
became the international representative of a small -arms 
manufactory, and spent fifteen years chiefly in travel 
through the several Spanish -American countries; and 
during this period he acquired an extended and intimate 
acquaintance with languages and peoples, as well as with 
national leaders and policies. Impressed by the oppor- 
tunities for international business presented by the 
actual and prospective republics of Spanish America, he 
established a house in Chicago, only to be ruined by the 
fire of 1871; later he combined business enterprises in 
St Louis and Mississippi City with notable researches 
in the archeology of the Mississippi valley. Unhappily 
pursued by conflagrations, he turned to research and 
publication, making important contributions to the 
projectors of the Pan-American Railway and the Bureau 
of American Republics. 


Colonel Hilder acted as secretary of the National Geo- 
graphic Society during the year ending June, 1899, after- 
ward becoming Ethnologic Translator in the Bureau of 
American Ethnology. He continued in this position to 
the time of his death, though he was detailed as a special 
agent of the Pan-American Exposition for work in the 
Philippines during the earlier half of 1900. Despite the 
briefness of his connection with the Bureau, he had 
already made himself a place among the most valued and 
trusted members of the corps. 

As indicated by his career, Colonel Hilder possessed 
remarkably strong character ; yet he was by instinct a 
naturalist and student, and devoted the best energies of 
his life to the increase and diffusion of knowledge. His 
later publications, through the Bureau of Education and 
the Bureau of American Republics, as well as through the 
National Geographic Magazine, the Forum, and other 
standard periodicals, are well known ; while his graceful 
and instructive lectures, based on personal observations 
in India, Egypt, South Africa, ' Central America, the 
Philippines, and other remote regions, live in the memory 
of thousands. 


Appropriation by Congress for the fiscal year ending June 30, 1901, "for 
continuing ethnological researches among the American Indians under 
the direction of the Smithsonian Institution, including salaries or com- 
pensation of all necessary employees and the purchase of necessary 
books and periodicals, fifty thousand dollars, of which sum not exceed- 
ing one thousand five hundred dollars may be used for rent of build- 
ings" (sundry civil act, June 6, 1900) $50, 000. 00 

Salaries or compensation of employees $34, 080. 45 

Special services $526. 35 

Traveling and field expenses 2, 1 12. 82 

Ethnologic specimens 3, 388. 78 

Manuscripts 2, Oil. 00 

Drawings and illustrations 407. 95 

Negatives 10. 40 

Books and periodicals for library 822. 58 

Office rental 1, 500. 00 

Office furniture 683. 33 

Lighting 94. 53 


Stationery, supplies, etc $1, 238. 04 

Freight 257. 93 

Postage, telegrams, etc 72. 50 

Miscellaneous 108. 65 

$13, 234. 86 

Total disbursement . . , $47, 315. 31 

Balance July 1, 1902 2, 684. 69 


The three papers presented in connection with this 
Report mark a forward step in the systematic researches 
undertaken by the Bureau, and relate to fields already 
more or less fully occupied by the authors represented. 
The paper of Dr Fewkes, Two Summers' Work in Pueblo 
Ruins, is based on explorations made in the heart of the 
pueblo country during the years 1896 and 1897, and 
serves, in connection with his previous writings, to extend 
our knowledge of the local tribes back into the shadowy 
time that witnessed the gathering of the clans on the 
plateaus and in the canyons of the Colorado valley. The 
ruins described are located in the middle part of the 
valley of the Little Colorado and in the upper Gila valley, 
to the south. In the prosecution of these researches Dr 
Fewkes has made much progress in verifying traditions 
of the Hopis and in determining affinities and movements 
of the prehistoric communities; and, by utilizing his rich 
fund of accumulated knowledge, he has succeeded in 
contributing materially to our understanding of the unique 
culture of this remarkable region. 

The paper by Dr Thomas, Maya Calendar Systems, 
deals with the interpretation of the ancient records of the 
Maya tribes of Middle America. Not having discovered 
an American Rosetta stone, the students of this impor- 
tant branch of native culture have given chief attention 
to the more pregnable features of the records — the systems 
of numeration as applied to the native calendar. Largely 
through the persistent and most praiseworthy efforts of 
A. P. Maudslay, the known examples of glyphs, sculp- 
tured in stone and wood, modeled in stucco, and painted 


on walls, have been accurately recorded and published 
and are available to students the world over, and discus- 
sion can now proceed 1 to much better advantage than 
heretofore. A first paper on the Maya calendar systems 
by Dr Thomas appeared in the Nineteenth Annual Report 
of the Bureau, and the present paper continues the dis- 
cussion, with special reference to the interpretations of 
Mr Goodman. These papers necessarily assume a some- 
what controversial character, but this is probably not to 
be regretted, since the legitimate outcome of honest con- 
troversy is a closer approximation of the truth. Already 
some notable results have been achieved and are generally 
accepted, and a fair understanding of the entire calendar 
system may be looked forward to with confidence. 

"The Hako" is an important ceremony .enacted by 
various western tribes, and the record presented in the 
paper by Miss Fletcher is exceptionally full and satisfac- 
tory. This record was made among the Pawnees under 
most favorable conditions. After years of patient work 
the words and music of all the songs and explanations of 
the rites and the hidden meanings of every word and act 
were obtained. This was done mainly through the aid of 
an, old man of the Chaui band who is known as the Kurahus 
or leader of the ceremony. Like primitive ceremonies 
generally, this is a prayer, and its motives are peace, 
plenty, and abundance of offspring. It is intertribal, and 
not only serves as a means for the interchange of ideas 
through contact and through gifts but represents one of 
the many powerful agencies which, by spreading tolerance 
and friendly feeling, tend to weld scattered warlike bands 
of men into great peaceful nations. In its melodies, re- 
corded on the graphophone and transcribed into our 
notation by a skilful musician; in its meters, carefully 
studied and analyzed; and in the metrical translation, 
which has caught so perfectly the spirit of the original, 
there is abundant material for students of music, poetry, 
and religion. The Bureau is fortunate in being the means 
of presenting to the world this superb study of a typical 
aboriginal ceremony. 


22 ETH— <M 1 







Summer of 1896 17 

General outline 17 

Ruins on the Little Colorado river 20 

General features. . _• 20 

Ruins near Winslow 22 

The Homolobi group. _ . 23 

Location . . _ ". 23 

Former inhabitants 24 

Ruin 1 25 

Ruin 2 29 

Ruin 3 29 

Ruin 4 30 

The Chevlon ruin (Cakwabaiyaki) 30 

The Chaves pass ruin (Tciibkwitcalobi) 32 

Ruins between Winslow and the Hopi pueblos 34 

Cavate ruins near Flagstaff 35 

Old caves 36 

New caves 37 

Turkey Tank caves 38 

Ruins near the Black falls of the Little Colorado 39 

Location and previous exploration 39 

General features 40 

Group a . 42 

The Citadel 43 

Ruin a 43 

Ruins B, C, and D 44 

Ruins E and f 44 

Ruins a, n, I, and J_ 44 

Ruin k 45 

Ruin L 46 

Ruin M 46 

Group b 47 

Ruin a 47 

Ruin B 50 

Ruin c 50 


Summer of 1896 — continued. Page 
Ruins on the Little Colorado river— continued. 

Ruins between Win§low. and the Hopi pueblos — continued. 

Ruins near the Black falls of the Little Colorado — continued. 

Group c . . - 51 

Ruin a 51 

Ruin b 54 

Ruins near Honani's house at Burro spring : 55 

Objects from the Little Colorado ruins ' 56 

Pottery ■_ 56 

General features 56 

Classification by color and surface finish 58 

Coarse unpolished ware . 58 

Undecorated polished ware 58 

Decorated polished, ware 58 

Red and brown ware 58 

Yellow ware 59 

Black ware 59 

Black and white ware 59 

Red and black ware 60 

Red, black, and white ware 60 

White and green ware 61 

Classification by form 61 

Food bowls '..___ 61 

Vases and jars 61 

Ladles 63 

Canteens .. . 64 

' Cups 65 

Animal forms 66 

Slipper-shaped vessels 69 

Decorative designs 69 

General character , 69 

Human figures 70 

Quadruped figures 71 

Bird figures '. 73 

Insect figures 81 

Arachnid figures 81 

Geometrical designs 83 

Ornaments 85 

Mosaics 85 

Lignite ornaments . . . 87 

Shell ornaments ' . . 88 

Bone implements 93 

Turtle carapaces - 95 

Horn objects : 96 

fewkes] CONTENTS 7 

Summer of 1896 — continued. Page 
Objects from the Little Colorado ruins — continued. 

Pigments - 96 

Cloth 97 

Matting 97 

Basketry 98 

Prayer-sticks 99 

Bow and arrows 100 

Gaming canes . 100 

Seeds 101 

Food 101 

Stone implements 102 

Stone slabs 104 

Disks i 106 

Fetishes 107 

Human crania 110 

Animal remains 110 

Miscellaneous objects 110 

Ruins of Old Shumopovi i 111 

General features 111 

Pottery from the ruins 113 

General features and form . 113 

Decoration 114 

Summer of 1897 120 

Introduction : _ 120 

Plan of the expedition 121 

Kintiel 124 

The ruin and its cemeteries "_ 124 

Pottery from the ruin 129 

General features and form 129 

Decoration 131 

Miscellaneous objects from the ruin 133 

Kinna Zinde 134 

Ruins near Holbrook 134 

Objects from Woodruff butte _ 135 

Ancient habitations in the petrified forest 135 

Four-mile ruin . . 136 

General plan 136 

A room in the ruin 137 

Suburban ovens _ _ . 139 

Cemeteries 139 

Collections 140 

Pottery 1 40 

Principal types 140 

Gila type 141 


Summer of 1897 — continued. Page 
Four-mile ruin — continued. 
Collections — continued. 
Pottery — continued. 

Forms. .. 142 

Decoration 142 

Human figures . 143 

Quadruped figures , 145 

Bird figures • 146 

Butterfly figures 148 

Feather decoration ... 149 

Geometrical figures 150 

External ornamentation on food bowls 152 

Rain-cloud designs . 155 

Character and treatment of mortuary pottery 158 

Potter's outfit 158 

Stone implements . 159 

Stone slabs 1 .... 160 

Copper bell _• 162 

Prayer-sticks , 163 

Gourd rattles 163 

Ornaments 163 

Bone implements '. 164 

Animal remains 164 

Ruins near Four-mile ruin ... 164 

Ruins at Pinedale _ . 164 

i The buildings .... 164 

Collections 166 

Stott ranch ruin 167 

Ruins in Pueblo Viejo . 168 

The valley and its history — general features of the ruins 168 

Distribution of ruins in Pueblo Viejo 170 

Epley's ruin 171 

Buena Vista 171 

Other ruins 173 

Changes in the valley since 1847 174 

Former population of the valley . . 175 

Cremation of the dead 175 

Architecture 176 

Terraced gardens 177 

Prehistoric irrigation in Pueblo Viejo 178 

Pottery from Pueblo Viejo 179 

Color and surface finish. . 179 

Undecorated rough ware 179 

Decorated rough ware 179 

fewkes] CONTENTS . 9 

Summer of 1897 — continued. Page 
Ruins in Pueblo Vie jo — continued. 

Pottery from Pueblo Vie jo — continued. 
. Color and surface finish — continued. 

Undecorated red ware 180 

Decorated black and white ware 180 

Decorated gray ware 180 

Decoration 180 

Stone objects from Pueblo Viejo 183 

Implements 183 

Stone slabs 185 

Other stone objects 186 

Shell ornaments from Pueblo Viejo 187 

Cliff houses on Bonita creek 187 

Ruins in the foothills of Mount Graham 187 

Sacrificial caves . 187 

Collections from a cave in the Nantacks '. 188 

Effigy vases from southern Arizona 189 

Distribution of decorated pottery in Arizona 192 

Evolution of the pueblo type of architecture _ 193 



Plate I. Old caves, near Flagstaff, Arizona 196 

II. New caves, near Flagstaff, Arizona 196 

III. Entrance to New caves, near Flagstaff, Arizona 196 

IV. Turkey Tank caves, near Flagstaff, Arizona _' 196 

V. Interior of cave near Flagstaff, Arizona 196 

VI. Interior of cave near Flagstaff, Arizona 196 

VII. View from the Citadel, group *a, Black falls, Little Colorado 

river, Arizona 196 

VIII. Ruin g. group a, Black falls, Little Colorado river, Arizona.. 
IX. Ruins h and J, group a, Black falls, Little Colorado river, 

Arizona 196 

X. Ruin J, group a, Black falls, Little Colorado river, Arizona 196 

XL Ruin J, group a, Black falls, Little Colorado river, Arizona 196 

XII. Ruin a, group b, Black falls, Little Colorado river, Arizona 196 

XIII. Section b. ruin a, group b, Black falls, Little Colorado river, 

Arizona 196 

XIV. Ruin a, group b, Black falls, Little Colorado river, Arizona 196 

XV. Chimney in ruin a, group b, Black falls, Little Colorado river, 

Arizona 196 

XVI. a, Reservoir, ruin a, group b, Black falls, Little Colorado 

river, Arizona; b, Small ruin near camp, Black falls, Little 

Colorado river, Arizona 196 

XVII. Ruin a, group c, Black falls, Little Colorado river, Arizona 

(view from the east) 196 

XVIII. Ruin a, group c, Black falls, Little Colorado river, Arizona 

(view from the west) . . 196 

XIX. Ruin a, group c, Black falls, Little Colorado river, Arizona 

(view from the north) 196 

XX. Vase from Chevlon ruin, Arizona 196 

XXI. Vase from Homolobi, Arizona 196 

XXII. Vases from Homolobi, Arizona ... 196 

XXIII. Food bowls from Homolobi, Arizona 196 

XXIV. Food bowl from Four-mile ruin, Arizona 196 

XXV. Food bowls from Four-mile ruin, Arizona. 196 

XXVI. Food bowl from Four-mile ruin, Arizona 196 

XXVII. Food bowls from Homolobi, Arizona 196 

XXVIII. Food bowls from Chaves pass and Chevlon, Arizona. 196 

XXIX. Food bowls from Chaves pass and Hornolobi, Arizona 196 




Plate XXX. Vases from Homolobi, Arizona . 196 

XXXI. Vases from Homolobi, Arizona 196 

XXXII. Vases from Homolobi, Arizona . 196 

XXXIII. Vases from Homolobi. Arizona 196 

XXXIV. Dippers from Chevlon, Homolobi, and Chaves pass, Arizona, 196 
XXXV. Food bowls from Chevlon, Homolobi, and Four-mile ruin, 

Arizona 196 

XXXVI. Pottery objects from Chaves pass and Homolobi, Arizona 196 

XXXVII. Food bowls from Chevlon, Arizona 196 

XXXVIII. Vases from Chevlon, Arizona 196 

XXXIX. Food bowls from Chevlon, Arizona 196 

XL. Food bowls from Four-mile ruin, Arizona 196 

XLI. Food bowls from Four-mile ruin, Arizona 196 

XLII. Food bowls from Four-mile ruin, Arizona 196 

XLIII. Food bowls from Chevlon and Homolobi, Arizona 196 

XLIV. Mosaic frog from Chaves pass, Arizona 196 

XLV. Bone implements from Chaves pass, Arizona _ . : ... 196 

XL VI. Stone slab with rain-cloud design, from Chevlon, Arizona:. 196 

XL VII. Vase and food bowl from Chevlon, Arizona 196 

XL VIII. Vases from Shumopovi, Arizona 1 196 

XLIX. Vase from Shumopovi, Arizona 196 

L. Vases from Shumopovi, Arizona , 196 

LI. Food bowls from Shumopovi, Arizona 196 

LII. Modern cemetery at Hopi pueblos, Arizona 196 

LILT. Plan of Kintiel ruin, Arizona 196 

LIV. Surface of mounds at Four-mile ruin, Arizona, before exca- 

vation 196 

LV. Food bowls from Epley's ruin, Gila valley, Arizona 196 

LVI. Food bowls from Four-mile ruin, Arizona 196 

LVII. Excavations at Four-mile ruin, Arizona 196 

LVIII. Plan of a room in Four-mile ruin, Arizona 196 

LIX. Vase and pitcher from Four-mile ruin. Arizona >, 196 

LX. Vase and food bowl from Four-mile ruin, Arizona 196 

LXI. Food bowls from Four-mile ruin, Arizona 196 

LXII. Vases from Four-mile ruin, Arizona 196 

LXIII. Food bowls from Four-mile ruin, Arizona 196 

LXIV. Spiral design on food bowl from Four-mile ruin, Arizona ___ 196 
LXV. Perforated stone slab and loom stones from Four-mile ruin, 

Arizona 196 

LXVI. Plan of Buena Vista ruin, upper Gila valley, Arizona 196 

LXVII. Food bowls and vase from Pueblo Vie jo, Gila valley, Arizona. 196 

LXVIII. Vase from Pueblo Viejo, Gila valley, Arizona 196 

LXIX. Food bowl and vases from Pueblo Viejo, Gila valley, Ari- 
zona 196 

LXX. Map showing distribution of ancient pottery in Arizona 196 




Figure 1. 



I . 








Map of Ancient Tusayan 21 

Ruin 1 , Homolobi 26 

Plan of an ' ' Old cave ' ' dwelling 37 

The Citadel, group a 43 

Plan of ruin J, group a 45 

Section a, ruin a. group b . _ 46 

Plan of section a, ruin a, group b 47 

Plan of section b, ruin a, group b 48 

Ruin a, group C, from the south 51 

Plan of ruin a, group c 52 

Plan of ruin b, group c 54 

Ladle with divided handle, from Chevlon 62 

Ladle with figure on handle, from Chevlon. 63 

Cup with bird designs 64 

Dipper from Homolobi 65 

Mug from Homolobi 65 

Cup, rough ware, decorated, from Chevlon 65 

Mug . . 66 

Mug from Chevlon 66 

Duck-shaped vessel from Chevlon 67 

Jar with four knobs, from Homolobi 67 

Bird-shaped vessel from Chevlon 68 

Bird-snake vase from Chevlon . _. 68 

Footprints on inside of vase from Homolobi 70 

Quadruped figure on food bowl from Chaves pass 71 

Quadruped figure on food bowl from Chevlon 72 

Vase with bear's-paw design 73 

Mythic bird figures and rain-cloud symbols on food bowl from 

Chevlon 74 

Vase with four bird figures, from Homolobi 75 

Mythic 1 ird figure on food bowl from Chaves pass 75 

Bird design on food bowl from Homolobi 76 

Bird figure on food bowl from Homolobi 77 

Mythic bird figure on food bowl from Chevlon ». _ . 78 

Bird figure on food bowl from Chevlon 79 

Food bowl with bird designs 80 

Vase with bird symbols, from Homolobi .... 81 

Spider and sun emblem on food bowl from Homolobi 82 

Three lines of life; design on food bowl from Chevlon 83 

Geometrical designs on food bowl from Chaves pass 84 

Broken fret on food bowl from Chevlon 85 

Food 1 )owl with geometrical patterns 85 

Mosaic gorget from Chaves pass 86 

Bone ear pendants from Chevlon 86 

Lignite ear pendant from Chevlon 87 



Figure 45. Lignite gorget 88 

46. Incised armlet from Chevlon 89 

47. Armlet with inlaid turquoise, from Chevlon 90 

48. Shell used for rattle, from Chevlon 91 

49. Shell ornament from Homolobi . . 92 

50. Shell frog from Chevlon 92 

51. Shell object from Chevlon 92 

52. Shell gorget from Chevlon 92 

53. Bone awl from Chaves pass '. 94 

54. Carved bone awl from Homolobi 94 

55. Bone implement from Chaves pass 94 

56. Bone tube from Homolobi .... 95 

57. Stick used by stick swallower . from Chevlon 95 

58. Disk of turtle shell, from Chevlon 96 

59. Kaolin cup from Chaves pass ... 96 

60. Matting from Chevlon 97 

61 . Basketry of Oraibi type, from Chevlon -. ,. _ _ _ 98 

62. Basketry of Oraibi type, showing manner of plaiting , 99 

63. Basketry of coiled type, from Chevlon 99 

64. Gaming canes from Chevlon -_ 101 

65. Stone implement from Homolobi 102 

66. Stone ax from Homolobi 102 

67. Copper bell from Chaves pass Ill 

68. Bowl with snout, from Shumopovi 114 

69. Plumed snake; design on food bowl from Shumopovi 114 

70. Mythic bird and game of chance; design on food bowl from 
Shumopovi 115 

71. Bird design on food bowl from Shumopovi 116 

72. Mythic bird design on food bowl from Shumopovi 117 

73. Symbolic bird design on food bowl from Shumopovi 117 

74. Gambling canes and bird; design on food bowl from Shumopovi. 118 

75. Dipper with decorated handle, from Kintiel . 130 

76. Coiled vase from Kintiel ... 130 

77. Two-handled bowl from Kintiel 131 

78. Globular bowl from Kintiel 131 

79. Handle of dipper from Kintiel 131 

80. Frog design on bowl from Kintiel 132 

81. Food bowl from Kintiel 132 

82. Bird design on food bowl from Kintiel 132 

83. Cup from Kintiel 133 

84. Dipper from Kintiel 133 

85. Stone birds from Woodruff butte 135 

86. View of Four-mile ruin from river bed 136 

87. Upright posts in wall at Four-mile ruin 138 




Figure 88. Ornamented rough bowl from Four-mile ruin 141 

89. Small saucer from Four-mile ruin 142 

90. Human figure on food bowl from Four-mile ruin 144 

91. Bird design on food bowl from Four-mile ruin 146 

92. Bird design on food bowl from Four-mile ruin 147 

93. Bird designs on food bowl from Four-mile ruin 147 

94. Bird design on food bowl from Four-mile ruin 148 

95. Butterfly design on food bowl from Four-mile ruin 149 

96. Sun emblem on food bowl from Four-mile ruin . 150 

97. Bowl with double spiral design, from Four-mile ruin 151 

98. Decorated vase from Four-mile ruin 151 

99. Unknown design on food bowl from Four-mile ruin 152 

100. Bear design on exterior of food bowl from Four-mile ruin 153 

101. Twin bird design on exterior of food bowl from Pinedale 154 

102. Bird design on exterior of food bowl from Four-mile ruin 154 

103. Bear's-paw design on exterior of food bowl from Four-mile 

ruin i 155 

104. Geometrical design on exterior of food bowl from Four-mile 

ruin 155 

105. Cloud emblem on food bowl from Four-mile ruin 157 

106. Stone used in belt frame, from Four-mile ruin 160 

107. Stone slab from Four-mile ruin 161 

108. Copper bell from Four-mile ruin . 162 

109. Bone implement from Pinedale ruin 166 

110. Decorated slipper jar from Pueblo Viejo 181 

111. Moccasin-shaped jar from Pueblo Viejo 181 

112. Arrow polisher from Pueblo Viejo _ 182 

113. Arrow polisher from Pueblo Viejo 183 

114. Metate from Pueblo Viejo 184 

115. Unknown stone object from Pueblo Viejo 184 

116. Ceremonial stone slab from Pueblo Viejo 185 

117. Ceremonial stone slab from Pueblo Viejo 186 

118. Stone fetish from Pueblo Viejo 186 

119. Indented bowl from a cave in the Nantacks 188 

120. Small amphora from a cave in the Nantacks 189 

121. Human effigy vase from a cave in the Nantacks 189 

122. Effigy vase from Pueblo Viejo 191 


By Jesse Walter Fewkes 



The following report embodies the results of archeological field work 
for the Bureau of American Ethnology in the summer of 1896. a 

The author was fortunate in having as his assistant Dr Walter 
Hough, of the National Museum, who, at his suggestion, took up in 
addition a study of Hopi ethnobotany, the results of which study have 
already been published. 6 

A week after his departure from Washington on May 30 the author 
began excavations at a ruin called Homolobi, near Winslow, Arizona, 
where he worked continuously until the close of June. During July 
and a part of August he excavated ruins at the mouth of Chevlon fork, 
on the Little Colorado river, and at Chaves pass. c 

The short distance of Homolobi and the Chevlon ruin from Winslow 
allowed him to make daily trips from that town to the ruins, where 
the workmen were encamped. At the close of each day's work the 
objects found on that day were carried to the hotel, where they 
were catalogued and packed for shipment. Even with this precau- 
tion some of the specimens were appropriated by visitors attracted by 
the beauty of certain of the pottery objects. While archeological work 
in the vicinity of a town lias advantages so far as the practical work 
of boxing and shipping are concerned, it has many disadvantages, one 
of the least of which is that just mentioned. 

The considerable distance of the Chaves pass ruins from a town 
necessitated a camp at the diggings, which was far from a hardship, 
considering the beautiful forests and the fine water near the ruins. 
A camp was made at Old Shumopovi during a short stay at this ruin. 

« A preliminary report on the field work of this year may be found in the Annual Report of 
the Smithsonian Institution for l89f>. 

& American Anthropologist, May, 1897. 

'This refK>rt was written and transmitted for publication in 1898. Since that time there has 
been considerable archeological activity in Arizona, and several collections have been made in 
the region, even in the ruins here considered. Certain specimens in these collections add impor- 
tant data to the discussion of the culture and migrations of the people who once inhabited this 
part of the Territory. 

22 ETH— 04 '2 17 


Tlie objective material collected numbered 1,875 entries in the 
National Museum catalog, but since many of these entries include sev- 
eral objects, the actual number of specimens obtained was somewhat 
larger. The specimens were collected from the following localities: 

Homolobi 700 

Chevlon fork (Hopi name, Cakwabaiyaki) 635 

Chaves pass (Hopi name, Tcubkwitcalobi) 284 

Shumopovi ■'_ J i 108 

Various other ruins 97 

« 1,824 

Other specimens were obtained from various ruins on the Little 
Colorado, and from Mishongnovi, Awatobi, Sikyatki, and Old Walpi. 

By far the greatest number of specimens collected were objects 
of a mortuary nature from the cemeteries. Although many of these 
were broken in getting them out of the ground, it is estimated that 
over one-half were entire, and fully two-thirds of the remainder 
have been so well repaired that they answer all the purposes of the 
student. The breakage was in part due to the inexperience of the 
workmen, but most of the vases, bowls, and the like were fractured 
by the earth, logs, or stones thrown on them in the graves when they 
were buried. Each bowl appeared to have formerly contained mor- 
tuary offerings of some kind, as food, paints, or prayer-sticks, and it 
was not rare to find food bowls piled up in nests one within another. 
There is no conclusive evidence that any large number of vessels 
were broken by design when interred. 

At the close of the archeological work on the ruins mentioned, the 
author witnessed the Flute ceremony at Mishongnovi, and the Snake 
dances at Oraibi, Shipaulovi, and Shumopovi. Some of his observa- 
tions on the latter have been published in the Sixteenth Annual 
Report of the Bureau of American Ethnology, and a short description 
of the Mishongnovi Flute altars, observed in 1896, appeared in the 
Journal of American Folk-Lore. The author also made a few studies 
of the Walpi Flute observance, which supplemented those already 
published elsewhere, and enabled him to prepare an extended memoir 
on this important ceremonial as performed on the East mesa. 6 

On his return to the railroad, after the close of the summer's work, 
the author visited Zuni to prospect for ruins in anticipation of future 
exploration, and made a flying trip to the pueblos Isleta, Sandia, and 
Tesuki. A small collection of ethnological objects was made at these 
pueblos, and other specimens were purchased at Santa Fe; these, con- 
sisting of old paintings on skin from ancient pueblo missions, have 
been presented to the National Museum. It is believed that there is 
considerable material in the hands of traders or others in the South- 
west, illustrative of the early mission period, which ought to find a 

"Besides the 1,824 specimens catalogued in the field 51 additional objects were entered in the 
National Museum catalog. 
&See Nineteenth Annual Report of the Bureau of American Ethnology, 19(X), part 2, p. 952-1011. 


permanent home in the National Museum. Some of these objects are 
very old, and show a mingling of Pneblo and Christian symbolism 
which is highly instructive, but in the rapid extinction of old manners 
and customs they are being replaced by more modern objects, and 
will soon disappear completely. Their preservation might well occupy 
the attention not only of the ethnologist but also of the historian. 

Although the visit to Pueblo Sandia was a short one, of the nature 
of a reconnaissance, it was full of interest. For some unknown reason 
this pueblo seems to have been overlooked by most ethnological 
students of the pueblos, but to one interested in the Hopi Indians, 
Sandia presents many highly instructive problems. It is peopled by 
descendants of the people of Payiipki, now a ruin on the Middle 
mesa, and no doubt the Sandians have legends of the former home of 
their ancestors in Tusayan/ 1 

Sandia has a large kiva, not unlike those in other Rio Grande 
pueblos, where old rites are undoubtedly still perpetuated. It would 
be interesting to know something of the nature of these ceremonies, 
in order to compare them with those of the Tusayan ritual. 

The author hoped that he would be able to find some ruins in the 
immediate neighborhood of Holbrook, Arizona, and he visited the 
mesa north of the town with that thought in mind; but he was 
disappointed, although evidences of temporary camps and a fewpicto- 
graphs were discovered. He heard, however, of ruins at Carrizo, and 
saw a few beautiful specimens of stone obj-ects from that locality. 
The trip from Navajo Springs to Zuiii failed to reveal any consider- 
able ruins along that trail, but the examination was a superficial one. 
There are several large ruins not far from the Navajo railroad station, 
which were not examined. 6 

There are small ruins on some of the hills of the bad lands of the pet- 
rified forests c near Holbrook. One or two of these are of consider- 
able size, and many objects indicative of former visitants or occupants 
were found on the ground about them. The author succeeded in dis- 
covering a single grave in one of the mounds, and excavated from it 
a few fragments of pottery, but these objects did not occur in suffi- 
cient quantities to justify extended work. Not far from Adamana 
station, on the Atlantic and Pacific railroad, there is a large ruin on 
a hill, which visitors to the petrified forest have no doubt noticed. 
This ruin is of considerable size, and promises a rich yield of 
archeological material should reasonable excavations be made out- 
side its walls. 

"On a map by Menchero the site of Payiipki is called •• Mesa de las TiKuas.*' indicating that it 
was peopled by Tiwas. Some of the Hopis Stay that relatives of the Asa clan once lived there. 

b Some of these ruins were specially studied in the summer of 1W7. and will be described later 
in an account of the operations of that year. 

'This remarkable collection of fossil trees is about 15 miles from Holbrook, and may be called 
one of the wonders of Arizona. There are in reality three petrified forests, or three places where 
the bad lands are eroded sufficiently to lay bare the huge fossil trees which they cover. The 
signs of former habitation observed in the section nearest Holbrook show comparatively late 



General Features 

The plains and mesas bordering the Little Colorado river and its 
tributaries were sites of populous pueblos in prehistoric times. There 
remain many descendants of this former population who now inhabit 
pueblos distant from that stream. The Zunis alone still live on the 
bank of one of its tributaries, and from the source of the river 
to its mouth the ancient pueblos have long since been deserted. It 
is asserted by certain Tusayan clans that their ancestors "formerly 
inhabited the pueblos now in ruins on this river, and traditionists 
have names for these villages. The plan of the present expedition 
was to explore ancient ruins claimed by the traditionists of the Patki, 
Patnii, and Piba people as a former home of their families, in order to 
determine the truih of their legends and to gather what archeological 
data there were bearing on the prehistoric migrations of the people 
who inhabited the western section of the pueblo area. 

The ruins along the Little Colorado do not differ greatly in general 
character from those in the vicinity of the inhabited Hopi pueblos. 
They are situated both in the river vallej 7 and on bordering mesas, 
and, owing to the open character of the country, are mostly of the vil- 
lage type. Some.of the tributaries of the Little Colorado in the Mogo- 
llon mountains are said to be overlooked by cliff houses, several of 
which, in Clear creek canyon, still remain well preserved, according 
to report; but these ruins have not been investigated. 

Drifting sand has buried most of the ruins of the valley so deeply 
that the walls of few of them remain standing above ground. As a 
rule they are built on natural mounds, which, near Chevlon fork, have 
a gravelly character. 

There is little to guide one in a determination of the probable age 
of the ruins. No evidences of Spanish influence were detected in the 
excavations, but this does not, of course, necessarily mean that the 
pueblos were not inhabited contemporaneously with, or long after, 
the advent of the Spaniards. 

It is instructive to determine the probable causes of the evacuation 
of these river villages by ancestral Hopi clans. Among other influ- 
ences, the following may be mentioned. In the legends concerning the 
forays of the Apaches it is always recounted that they attacked the 
Hopi pueblos from the south. Although these vigorous nomadic peo- 
ple originally came from the north, they seem to have early taken 
possession of the portion of Arizona between modern Tusayan and 
the southern boundaries of the Territory, raiding as they wished the 
Pima settlements on the south, and the Hopi on the north. The 
exposed pueblos along the Little Colorado were poorly adapted for 
defense, and this may have led to their abandonment. 




Some years ago two Mormon towns were built not far from the pres- 
ent site of Winslow, and contiguous to Homolobi. These towns, Brig- 
ham and Sunset, were prosperous for many years, and their inhabitants 
cultivated extensive farms, which were irrigated from tributaries of 
the Little Colorado. The remains of one of their acequias can still 
be seen skirting the river side of Homolobi, and many of the stones for 
the walls of the towns are said to have been obtained from the Indian 
pueblo. The Mormon town is itself now a picturesque ruin, having 


/«. Mik.. 

•7b, Bute nAmm \ ..;*8-. 


Ruins Q/f fffitfoms Canyon j. 
Shlnno-pS*'" . •Sh.-fHtuyiovf-- - ~_ i-vOc^ 

"Giants Chair a Spr JJe/fy/oSpr 


Fk;. 1. Map of Ancient Tusayan. 

i Itinerary indicated by dotted lines.) 

been gradually abandoned. One reason for the desertion of Sunset 
is said to have been the alkalinity of the soil, which irrigation had 
developed. If this explanation accounts for the failure of the Mor- 
mon farmers, it might also apply to their I Topi predecessors. The 
failure of crops may have Led the Indians to seek other localities 
better suited for farming. However thai maybe, at the present time, 
189G, the river valley opposite Homolobi has been turned into a profit- 
able farm by a Winslow farmer, and when the author worked at 


Homolobi this farm was green with alfalfa and various market vege- 

A failure of the rain and the corn crop is distinctly mentioned as 
one of the causes which led the Patki and other southern Hopi clans 
to leave their settlements along the Little Colorado, but it is also 
stated that they were afflicted by a kind of gnat or sand flea in some 
of their earlier halting places. Possibly their dwellings became so 
infected with vermin as to lead to their abandonment. a 

The Little Colorado river was dry during the work at Homolobi, and 
was crossed and recrossed almost anywhere, the sole obstruction being 
the steep banks, which were several feet high. Late in the summer, 
however, it became a raging torrent, impassable save in one or two 
places, and even these were dangerous on account of the man}^ quick- 
sands. It is not improbable that the great freshets of the river may 
have had an important influence in the abandonment of the second 
ruin of the Homolobi group, one side of which is completely worn 
away, although of course it is not unlikely that this happened after 
its abandonment. Evidence of similar erosion is also apparent on the 
river side of ruin 1 of the Homolobi group; cemeteries on that side, 
if they ever existed, have long since disappeared. 

Ruins near Winslow 

As has been noted, the Hopis say that the ancestors of the Patki 
or Water-house 6 people lived in the far south. This tradition is very 
definite, and it even declares that they once inhabited a pueblo 
called Homolobi, stating that the position of this ancestral dwelling 
was near where -the railroad crosses the Little Colorado, not far from 
the town called Winslow. With this exact statement as a guide the 
author went immediately to that town, having made arrangements 
with some Hopi workmen to join him there. 

The ruins on the Little Colorado near Winslow were identified as 
the Homolobi group by Mr Cosmos Mindeleff, who mentions 10 ruined 
pueblos in this immediate vicinity, but his reference to them is brief, 
and includes no attempt at description. The best-informed inhab- 
itants of Winslow were ignorant of the existence of ruins near their 
village, and but for the confidence he had in the legends of the Hopis 
the author also would have doubted their existence. 

The site of Homolobi was found to be exactly where the Hopi stories 
placed it, and archeological results confirmed the identification. The 
author found not only Homolobi, but also three other ruins in the imme- 
diate neighborhood of Winslow, and before a month had passed demon- 


"The trail from Beaver Head in the Verde valley to Flagstaff, especially not far from. Rattle- 
snake Tanks, has a very bad reputation for the small gnat, which gives much annoyance to 

& The name Water-house means cloud, and the members of this clan are called both the rain 
and the cloud people. 


strated that this was one of the richest fields in Arizona for archeo- 
logieal work, although previous to this visit not a single specimen had 
been described from the region. 

It was also the author's desire to see how the ruins of the Little 
Colorado south of Tusayan were connected with those on the banks 
of its tributary, the Zuni river, higher up the watershed. For that 
purpose he examined somewhat in detail a ruin opposite the station 
Hard}', on the Atlantic and Pacific railroad, near where Chevlon 
creek empties into the Little Colorado river. This ruin will be called 
in this report the Chevlon ruin. Its Hopi name is Cakwabaiyaki. 

The objects exhumed from the cemeteries of the Chevlon ruin have 
close likenesses to those of ancient Zuni ruins, as well as to those of 
Homolobi, which is claimed by the Hopis. A logical interpretation of 
these resemblances would seem to be that the culture differentiation 
of the two peoples was not as wide in ancient times as it is at present. 
The inhabitants of the villages of the Little Colorado and its tribu- 
tary, the Zuni river, were formerly closely related, and, no doubt, 
when these villages were deserted, some of the clans went to Zuhi and 
others to the liopi pueblos. In subsequent times greater differentia- 
tion took place, which led to the present conditions. 

It was also desirable to push the examination of the ruins of this 
drainage area as far south as possible, for which reason two ruins in 
Chaves pass, about 30 miles south of Winslow, were investigated. 
This was the southern limit of field work in 1896, and in the last 
month of the summer the author followed the trail north to connect 
the Homolobi ruins with those of the Hopi reservation. 

We have good evidences from historical and legendary sources that 
there were inhabited pueblos between Zuhi and Awatobi as late as 
the middle of the seventeenth century. One of these, that of the 
Cipias (Tcipiya, according to the Ilopis), is distinctly mentioned as 
west and south of Zuni. It is not probable that all clans of the Patki 
people had wholly deserted Homolobi in the sixteenth century, and 
they may have been dwelling there as late as 1700. It is as yet an 
unanswered question whether any one of the ruins which were exca- 
vated in 1896 is Tcipiya, which, according to the Ilopis, the Zuiiis 
declare was midway between Awatobi and Zuhi. 


There are four ruins near Winslow, which may be called the Homo- 
lobi group and are provisionally numbered 1, 2, .'3, and 4. Of these, 
ruin 1, true Homolobi, yielded the best archeological results, and was 
nearest to the town, being about 3 miles away. More excavations 
were attempted at that place than at all the others. The ruin num- 
bered 2 is about three miles farther down the river and more distant 


from it, but is on the same side. Ruin 3 lies on the left bank of the 
river, about midway between the first and second, and ruin 4 is a few 
miles beyond on the same side, somewhat removed from the river. 
All of these ruins are thus within a radius of G miles of the town of 

Former Inhabitants 

Several Hopi clans, belonging to groups called the Water-house, the 
Squash, and the Rabbit, are said to have lived in these settlements 
along the Little Colorado, near Winslow. Among the clans of the 
first-named group may be mentioned the Corn, Agave, Rain-cloud, 
Lightning, and various others whose totems are aquatic animals. A 
list of them follows : 

Patki Water-house 

Kaii Corn 

Omauwii Rain-cloud 

Tanaka Rainbow 

Talawipiki Lightning 

Kwan Agave 

Sivwapi Bigelovia graveolens , 

Pawik Aquatic animals (Duck) 

Pakwa Frog 

Pavatiya Small aquatic animals, Tadpole (pakwa, frog; tiyo, young) 

The prominent chiefs of Walpi who belong to the Patki or Water- 
house people are Supela, Kwatcakwa, Sikwistiwa, and Kwaa. Ana- 
wita of Sichumovi is also a member of the family. The legends of 
Homolobi were told to the author by the last mentioned, but Supela 
and the others have much lore concerning this group of ruins which 
has never been published. 

The Patuii or Squash people, now extinct on the East mesa, are also 
said to have lived at Homolobi, and they are reputed to have settled 
at Tcukubi on the Middle mesa, and the Tawa or Sun people, who 
are associated with the Pakab or Reed clans, once lived with the 
Rain-cloud and Squash people in the Homolobi settlements. 

There is evidence from the present Hopi ritual that the Patki, 
Patuii, Tawa, and Piba (Rabbit) families, among others, lived at 
Homolobi. For instance, it is stated that the following clans intro- 
duced the societies and ceremonies mentioned, with their fetishes, 
into the modern system : 

Clans Ceremonies Societies 

p ,-, • (Soyaluiia . _1 (Kwakwantu 

iPalulukontiJ iLalakoiitu 

Piba New-fire Tataukyam u 

Patuii New-fire Wuwutcimtu 

This statement is supported by the facts that the chief of the Kwa- 
kwantu, the great warrior society," is Anawita, and that Supela is one 

" Kwahu, eagle; kwan, agave. 



of the chiefs in Soyaluna. The Sun priest, Kwatcakwa, takes a 
prominent part in the screen drama of Paliilukoiiti. The Kwakwantu, 
Eagle-agave people, are distinctly southern, coming from the region in 
Arizona where the great cactus or agave grows, and an examination 
of details of the ceremonies mentioned shows an instructive likeness 
to Mexican rituals. In both Soyaluna " and Paluliikonti the effigies of 
the Plumed Snake play important parts, and this conception is dis- 
tinctly a Mexican one, recalling Quetzalcoatl. It is for those cere- 
monials in which there is the closest likeness to Nahuatl rites that 
southern origin is claimed by the chiefs and other participators. 

Ruin 1 

As has been stated, the nearest of the Homolobi ruins to Winslow — 
the one which was chosen for extensive excavations — is about 3 miles 
distant on the right bank of the river, and about the same distance 
from where the river is bridged by the Atlantic and Pacific Railroad. 
At that point there are several small knolls separated by level spaces 
which might once have been cultivated fields. Between the site of the 
ruin and the river bank there is a small grove of cottonwood trees, 
the bases of which are evidently at times flooded by the overflow of 
the stream. 

The river takes a general northwest course from the railroad cross- 
ing, and is bordered with cottonwoods. On either side the banks are 
low and sandy, frequently caving in, rendering it treacherous to 
approach at certain seasons of the year. 

The neighboring plains are parched and dry in the spring and early 
summer, and violent sandstorms sweep over them, oftentimes so dense 
as to obscure all outlook. At these times work along archeological 
lines is very disagreeable, and life in the village is far from pleasant. 

When approached from Winslow the ruin is indicated by a number 
of low mounds without standing walls, and when it was first visited 
there Avas little to indicate that it was the site of a former pueblo, save 
many fragments of pottery strewn over the surface. The indications 
that Homolobi would be a profitable field for archeological investiga- 
tion were very small. 

The excavation of ruin 1 of the Homolobi group began about the 
first day of June, when a force of 5 Mexicans was employed to open 
the mounds at the northeast angle. The results of the work were not 
very satisfactory. They betrayed the fact that Homolobi was a pueblo 
of small size and of irregularly rectangular shape, with its highest walls 
on the northeast side. Considerable broken pottery, some stone 
implements, and other objects were obtained, but all the evidences 
appeared to indicate that the more valuable specimens were removed 
when the pueblo was abandoned. 

"An illustrated account of this winter solstice ceremony may be found in the American 

Anthropologist, v. 1 1, March and April, 1898. 



fETH. ANN. 22 

The workmen penetrated to the lower floor, and found that the 
pueblo was two stories high at this point. The rooms were large and 
the beams of the flooring were well preserved. The floors of the rooms 
were large, flat stones; the lower chambers were nicely paved. The 
walls were made of stone masonry, nicely plastered, and in some 
instances blackened by smoke. In one of the largest of these rooms 
the floor stones were in two cases found to be perforated by round 
holes about the size of a sipapu in modern kivas. These slabs are 
in many respects similar to those found in graves outside the walls of 
the pueblo. 

Two human skulls, one of which indicated an old person, and 
several human bones were found on the floor of chambers in the 
northeast part of the ruin, and were supposed to represent intramural 





/v Trees _ 

C? O 

Fig. 2. Ruin 1, Homolobi. 

burials. No pottery, however, was found in the vicinity of these 
skeletons, which fact would seem to indicate that they were not 
buried with customary mortuary offerings. 

Continued work on the side of the ruin toward the river revealed 
the fact that this part had been worn away by the overflow of the 
stream, and a section had been cut through it in digging an irrigating 
ditch which formerly supplied the plains around Sunset with water. 

The osteological collection from Homolobi was very large. Early 
in his excavations the author was surprised at the number of animal 
bones which were thrown out by the workmen, especially after they 
had penetrated to some distance below the surface. There appears 
no better explanation for the existence of these bones than that they 
were remains of animals domesticated or used as food. These bones 
were carefully gathered, and have been identified by Mr F. A. Lucas, of 
the National Museum; a complete list of species found at Homolobi 
is published in this report, page 110. 


The occurrence of a skull of the domesticated dog in one of the 
graves at the Chaves pass ruin is significant, showing that this ani- 
mal was known to the ancients, and probably utilized by them. The 
fact that this dog was the broad-faced variety is particularly instruc- 
tive. It was not apparently a domesticated coyote or a mongrel like 
those which now are so common in some of the pueblos. Mr Lucas 
has published the following account a of this specimen: 

Among the many objects obtained by Dr Fewkes last summer from the ruined 
pueblo of Chaves pass, Arizona, is the cranium of a domesticated dog found in a 
grave with a human skeleton. Although the mere fact of a dog being discovered 
under such circumstances is in itself interesting, it is not at first sight remarkable, 
since it is well known that in America, as elsewhere, the dog was domesticated at an 
early date, and Clavijero mentions an ancient dog, which he calls " a quadruped 
of the country of Cibola, similar in form to a mastiff, which the Indians employ 
to carry burdens." Aside from the fact that this is the first dog's cranium dis- 
covered by Dr Fewkes, there are some points of special interest in the present 
case. Most of the Indian dogs are more or less wolfish in their aspect and have 
long skulls, with comparatively low foreheads, thus showing a small degree of 
specialization in the way of breed, and this is true of such of the mummied dogs 
of Egypt as I have seen. The cranium of the Chaves dog. on the contrary, is of 
the broad-faced type, with high forehead, and, curiously enough, is precisely simi- 
lar in size and proportions to the cranium of an Eskimo dog from Cumberland 
sound, the resemblance extending to the peculiar concavity and squareness of the 
nasal region. While this is an interesting coincidence, it is not brought forward 
as implying community of origin, but as instancing long domestication in order that 
so well-marked a breed could be established. A curious confirmation of the early 
origin of this breed was received from San Marcos, Texas, where, in excavating 
for ponds at the station of the United States Fish Commission, a human skeleton 
and bones of other animals were found in a layer containing many flint imple- 
ments. < >verlaid by two feet of black soil. The bones were those of existing species, 
including teeth of several bison, and there was also a fragment of a dog's skull 
similar in size and proportions to that obtained at Chaves pass. Owing to the 
circumstances under which the bones were exhumed, it is not known whether or 
not the dog and man were found together. While none of the bones were min- 
eralized, the condition under which they were found and the character of the 
human cranium showed them to be of very considerable age. 

Dr Fewkes states that the skulls of carnivores are used in Hopi religious cere- 
monies, and that the skull, paws, etc., are regarded as powerful fetishes of war- 
riors and cherished by them with much care. It is customary to bury a priest's 
fetishes with him. and there is little doubt that the dog's cranium from Chaves pass 
was a fetish of the man in whose grave it was found. As Dr Fewkes believes 
that the people of the Chaves pass ruin formerly lived far south, in contact with 
Nahuatl peoples, it can readily be seen how a dog's skull came to be part of the 
ceremonial outfit of the priest in whose grave it was found. 

The numerous turkey bones which were found do not necessarily 
mean that this bird was used as food by the ancient sedentary peo- 
ples of Arizona. We are told by the historian of Coronado's expedi- 
tion that the Pueblos bad domesticated fowls, but these were probably 
turkeys from which, as at present, were obtained feathers used in 

a Science, a. s.. v. 5, L897, p. 544. 


ceremonial practices. It would hardly seem possible that birds 
whose feathers were thus used would be eaten, although parallels 
to such a usage occur in the religious rites of many peoples. We 
know that their bones were made into needles and bodkins, and there 
is every probability that the wild turkey's flesh was eaten. a 

Because of the poor results of the early excavations at Homolobi 
the author determined to abandon the work at this ruin. A search 
for the burial places was not successful, although he carefully 
examined the sandy hillocks a few hundred feet away, expecting that 
they would be found there if anywhere. Meanwhile a half dozen 
Hopi Indians who had been sent for came to Winslow, and the author 
set them at work, having discharged the Mexicans. On the first day 
they discovered the cemeteries of Homolobi, after which it was only 
a question of time before a large collection was obtained. 6 

The burial places at Homolobi were close to the outer wall of the 
pueblo, so near to it in fact that the skeletons in some instances 
touched the outer face of the wall. These places of burial were liter- 
ally "under the ladders " of the town, if we believe, as we legitimately 
may, that the inhabitants formerly mounted to the house -tops by 

The outer slope of the mound was thus found to be crowded with 
the dead, and with them were multitudes of mortuary pottery offer- 
ings of all kinds. These cemeteries were found on the east/ side, 
opposite the river bank, and although it is possible that in ancient 
times burials may have been made on the side of the mound toward 
the river, if thev ever were there the overflow of the stream has 
washed them away or covered them up. 

The burial places were sometimes recognized by flat stones set on 
edge projecting above the surface of the soil. These stones had 
often fallen in over the grave, and were sometimes buried many feet 
below the surface. In many of these there was a small, round hole 
about the size of a broom handle; in others this hole was large and 
square. In one or two instances traces of pigment were detected on 
these burial stones, but in the majority the figures Avere not legible. 
The reason assigned by some of the Hopi workmen for these perfora- 
tions was for the escape of the soul; others regarded them as sym- 
bols of the sipapu, openings in the kiva floors which they closely 
resemble. In the pavement of one of the larger rooms which was exca- 
vated two slabs were found, one of which was perforated with a single 

a The eagles which are kept in the Moki towns to-day for the feathers used in ceremonies are 
buried at death in certain cemeteries with ceremonies of a simple character. 

*'In the winter of 1897 Dr G. A. Dorsey, of the Field Columbian Museum, made some archeo- 
logical studies of Homolobi, and he informs the author that he found many interments with 
beautiful mortuary pottery not far from the trenches made in 1896. Others have dug many 
specimens from the same ruin. 

<• While as a general thing the cemeteries to the east of a ruin are the largest, places of burial 
are not confined to that side of a ruin. 


hole, and the other, which was broken, had a hole in the middle and 
a round notch on the broken edge, as though there had once been a 
perforation at that place. 

Ruin 2 

Ruin 2 a of the Homolobi group, unlike ruin 1, is situated on top 
of a hill with a wide outlook. This is a much larger ruin than 1, and 
the Avails standing above ground are in a better state of preservation. 
No very extensive excavations were undertaken in this ruin, but a 
few graves were found some distance from the walls near the foot of 
the hill on which the pueblo was built. Several graves were indi- 
cated by upright slabs of stone set on edge in the soil, and from them 
a dozen or more mortuary vessels were excavated. The pottery was 
like that of ruin 1 in general character, yellow and brown ware pre- 

From the great size of this ruin the author suspects that if it were 
carefully excavated a rich collection might be found, but work upon 
it would be difficult, as it is situated quite a distance from water, 
and there are other practical difficulties, some of which, however, 
might be easily overcome. 

A number of bowls were found on the sides of the mesa on which 
this ruin stands, but these appear to indicate isolated burials; the 
cemetery was not discovered, and consequently the number of mor- 
tuary objects from the ruin Avas small. 

Ruin 3 

Ruin 3 of the Homolobi group is A^ery small, and is situated so near 
the present bed of the river that a portion of it has been worn away 
by the water. 

One of the most interesting features of this ruin is the use of blocks 
of adobe instead of stone in some of the partitions of the rooms. The 
situation of this ancient dwelling Avas such that stone Avas not easily 
obtained, and consequently, as so often happens elsewhere in the 
Southwest, adobe Avas utilized as a building material. 

Farther down the Little Colorado the author found in the ruins on 
the plains Avhich border the riA'er indications that the ancient houses 
were made of adobe alone, a fact readily explained by the absence of 
suitable stone on the site of the habitations. 

No other ancient Tusayan pueblo where adobe was used for the 
construction of houses is known, and for the most part to-day the 
building material is rock from the formation most convenient to 
the pueblo. 6 

'"This ruin, like ruin 1, was called Homolobi, and it is probable that the name is applied at the 
present time to the whole (duster of ruins near Winslow. As different phratries are reputed 
to have lived in this neighborhood, it may be possible to connect the several ruins with indi- 
vidual families. 

b There are adobe walls built out from the old mission at Awatobi. 


Very little in the line of exploration of ruin 3 was attempted, as only 
a. single visit was made to it. The pottery fragments were identical 
in character with those from the other Homolobi ruins, and the size 
of the mound shows that it was a small settlement. The stream has 
encroached on the foundations of the ancient town to such an extent 
that the cemeteries on this side have been obliterated. The sur- 
rounding plain was evidently cultivated, for remnants of old ditches 
can be seen in the neighborhood, though they were difficult to trace 
on account of frequent changes in the neighboring stream. 

One of the most interesting and exceptional objects taken from this 
ruin was a bone implement apparently made from the leg bone of 
an antelope. The blunt end of this object was carved in imitation 
of an animal, possibly a bear, the head, body, and legs being well 

Ruin 4 

Ruin 4 of the Homolobi group is one of the most picturesque in this 
region, and has many resemblances to Shipaulovi, on the Middle 
mesa of Tusayan. It resembles a castle perched on the pinnacle of 
a butte, which is steep on all sides. Its height is possibly 100 feet 
from the plain, and it has a wide outlook across the vallej^ of the 
Little Colorado. The top of the mesa is small and appears to have 
been covered with house walls built of stone, fragments of which 
have fallen down the steep sides of the mesa. 

The general ground plan is roughly rectangular, apparently with a 
central court, and the indications are that the houses were not more 
than one or two stories high. 

The debris at the base of the cliffs is full of fragments of pottery 
resembling that of ruin 1, and here undoubtedly we must look for the 
cemeteries, as there is no sign of a burial place on the top of the 
mesa. Near the foot of the mesa, and half way up its sides, border- 
ing the rough trail by which one can now ascend to the former site of 
the pueblo, there are many large bowlders, most of which are covered 
with pietographs pecked in the surface of the rock. These picto- 
graphs closely resemble those found almost universally in the western 
section of the pueblo area, and bear every evidence of being very 
ancient. Many of them were almost illegible, possibly from age, 
while others were fresh, suggesting more recent work. There are no 
painted pietographs, suggestive of the Apaches or other Indian tribes. 

No excavations were attempted at ruin 4, and the author's visit there 
was a short one. Although the ruin is not quite as large as ruin 1, 
the abundance of pottery fragments gave promise that it would be a 
fruitful field for archeological studies. 


Chevlon creek is a small stream of water which empties into the 
Little Colorado from its left bank about opposite the station called 
Hardy, on the Atlantic and Pacific railroad. It is possibly 15 miles 

fewkes] THE CHEVLON RUIN 31 

east of Winslow, with which it communicates by a tolerably good 
wagon road. About 2 miles south of Hardy, near the Esperanza 
ranch, the creek makes a graceful curve, west of which there rises a 
low mound — the site of Cakwabaiyaki or Blue Running Water pueblo. 

The country at this point is barren and sandy, with gravelly hills, 
and with rock jutting above its surface. There are no trees and only a 
few scrubby bushes of characteristic Arizona vegetation. The banks 
of Chevlon creek at this point are low, and in places there are numerous 
sand dunes. There is always water in the bed of the stream, but in 
the dry season much of it is lost when it gets to the thirsty sands of 
the Little Colorado valley. It is not potable, however, and animals 
do not drink it eagerly. Fishes and turtles in considerable numbers 
inhabit this stream. 

The road from Winslow to the Chevlon ruin crosses the railroad in 
the suburbs of the town, foliowing the plain to Salt slough, a putrid 
water hole, by whose alkaline waters many animals have been poisoned. 
From there the road leads to Clear creek, a beautiful stream which 
has been dammed to supply water for the town. A fine bridge has 
been built over Clear creek, and the water at that point is very deep. 
The banks are high and can3 7 on-like, and the spot is one of the most 
picturesque near Winslow. From this stream our Indian workmen 
obtained many turtles, which they highly prize, and they make fre- 
quent pilgrimages to it from Walpi to get water to use in their cere- 
monials. Not far from the Clear creek bridge there are evidences of 
a former population, and the broken-down walls of houses crown 
some of the adjacent hillocks. There are likewise many ancient picto- 
graphs in this vicinity. Higher up Clear creek valley, where it is dry 
and is called Clear creek canyon — a place A r isited on the way to 
Sunset pass — there are many evidences of former human occupation 
and abundant pictographs, some of which are of considerable interest. 

There are likewise said to be mounds similar to those at Homolobi 
on the banks of both Chevlon and Clear creeks, and there is little 
doubl that this is true — at all events as regards the former stream. 
Portions of canyons along the upper course of Clear creek were 
examined and numerous pictographs were found on their walls. 
There were also evidences of former habitations. 

It must have been not far from Chevlon ruin where Sit greaves and 
his part} 7 camped on October 2, 1851, but in his narrative he does not 
mention the ruin, though the short notice of camp number 9 corre- 
sponds with the locality in other respects. He says (page 7): 

The river [Little Colorado] here receives a tributary known among trappers as 
Chevelon's fork, from one of that name who died upon its hanks from eating some 
poisonous root. Their confluence produces an intricate labyrinth of sloughs, in 
which we became involved and were forced to encamp, not finding an outlet until 
late in the day. In several places veins of fibrous gypsum (selenite) were seen, 
looking like the ice crystals that hurst open the ground in spring." 

a Report of an expedition down the Zuni and Colorado rivers, Senate Document 59, Thirty- 
second Congress, second session, lsr>:i. 


This description may well apply to the mouth of the Chevlon in 
the month of August, when heavj^ rains are common, but in June the 
Little Colorado was repeatedly crossed near this point with the great- 
est ease, there being only a small rivulet to ford. Later, however, the 
river became a raging torrent, as the author found in attempting to 
ford it on the trip to the Tnsayan villages in July. 

The burials at Chevlon resemble those at Homolobi, and are simi- 
larly situated with relation to the ruin. Like those of the first ruin 
of Homolobi, interments were found in the largest number just out- 
side the outer walls of the pueblo, and at different depths. The 
configuration of the site of the ruin naturally introduced some 
modifications in the character of the burials. The drifting sand 
has buried them somewhat deeper at Chevlon than at Homolobi. 

No evidence of the cremation of the dead was discovered in the 
Chevlon ruin, at Homolobi, or in the ruins at Chaves pass. The dead 
in these three ruins were as a rule extended at full length, and not, as 
at Awatobi, placed in a sitting position. 


Looking southward from Winslow one can see in the distance a 
high range of mountains which separates the valley of the Little Colo- 
rado river from that of the tributaries of the Salt and the Gila. This 
range is broken at one point by a pass through which, in old times, 
there was a trail used I)} 7 Indians in trading excursions and migra- 
tions. It is called Chaves pass, from an old Arizonan named Chaves 
who was killed by Apaches near by. A small wooden cross in the 
open plain at the entrance of the pass is said to mark his burial place, 
and there are many other unmarked graves of white men who have 
lost their lives in this neighborhood. 

By taking the road south from Winslow one passes over a hilly 
country continually rising, with Chevlon butte far to the left, and, 
skirting Clear creek, follows it to Sunset pass, which is clearly visi- 
ble from Winslow. Beyond Sunset pass, where dwarf cedars afford 
a refreshing change from the treeless wastes about W inslow, the road, 
still rising, enters a well-wooded country between Sunset and Chaves 
passes. The road now becomes rougher, rising rapidly, with tall 
pines on all sides, until it passes an old well near the remains of 
a deserted cabin. This well is situated in Chaves pass, and there the 
road divides, one division continuing to Mormon lake and Rattlesnake 
tanks, where there is said to be a ruin of considerable size, and ulti- 
mately to the Verde valley, the ruins in which are numerous and 
extensive, the other to the Tonto basin. 

Two ruins lie on the hills above the pass; one, the smaller, is the 
first approached on the right-hand side; the other is so placed as to 
force the traveler out of his way, the road winding about it. Both 


are elevated above the trail through the pass, and from their house 
tops the observer can look across the valley, in which flows the Little 
Colorado, to the Ilopi buttes, far to the north. 

Their name, Tciibkwitcalobi, is derived from tciibio, antelope, 
kwitcala, notch, and obi, locative. The Navaho name Jettipehika has 
the same meaning. Both names were due to the fact that the pueblo 
lay in mountains where no short time ago antelope were abundant. 

During his stay in Winslow the author heard much about the ruins 
in Chaves pass and often gazed at the distant southern mountains, 
which particularly interested him as the possible gateway to Palatk- 
wabi, the Red land of the South. Chaves pass was fascinating in its 
archeological possibilities, for it was one of the few breaks in the 
rugged Mogollones through which ancient migrations could have been 
made. Accordingly, after examining the ruin at the mouth of Chev- 
lon fork, the author outfitted for a reconnoissance of the ruins which 
he expected to find in the pass. 

It need hardly be said that this was virgin ground for archeological 
work. No one, so far as is known, has ever mentioned these aboriginal 
habitations, which is not strange, considering the great number of 
undescribed ruins in this part of Arizona. Ruins at this point were 
especially interesting from the fact of their elevation and their posi- 
tion almost on the crest of the watershed of two great valleys, the 
Little Colorado and Gila, both of which were sites of large populations 
in prehistoric times. It is highly important to discover whether they 
furnish a connecting link between the two regions. There can be 
little doubt that the trail through the pass is an old one, and that it 
was used in the migrations of Indians. 

The two ruins at Chaves pass were built of the lava rock so abund- 
ant in this region. The larger must have been a pueblo of consider- 
able size, and covers an area much larger than any of the Homolobi 
group except ruin 2. The elevation on which it is built is consider- 
ably longer than wide, sloping abruptly, but is easy of access on all 
sides. The ruin is apparently of the rectangular type, with inclosed 
courts. It is composed of two house clusters connected by a range 
of rooms one and two series deep. Its rooms are square, and their 
outlines can be readily traced, though they are much obscured by 
fallen walls. In general type there is a close resemblance between 
the Chaves pass and Yerde valley ruins. 

Some attention was given to excavations in the rooms of both of 
the ruins, but the limited work there was not rewarded with great 
success. The walls of the rooms were built of blocks of lava rock, 
which had tumbled into the inclosures, and it was necessary to remove 
these before the floors were reached. Very little sand had drifted 
into the rooms on account of their elevated site, and the outlines of the 
rooms and the contours of the walls could be readily traced. No 

22 eth— Oi 3 


rooms could be identified as kivas, and the plastering was, as s\ rule, 
worn from the standing Avails. 

The burials at Chaves pass differ considerably from those at Homo- 
lobi and Chevlon. It was found that flat stones with perforations 
were not used in covering the burial cysts, but that logs — accessible 
in this wooded region — were employed. The bodies of the dead were 
extended at full length, and stones were laid at the head and feet. 
Upon these stones a number of logs were placed parallel with the 
bodies, and on either end of these logs there were other stones, 
generally bowlders, to hold them down. The body was thus covered 
with a rude floor, over which soil is now found. 

The depth of burial varied considerably, and it was not rare to 
find bodies 10 feet below the surface. The weight of soil above the 
logs had been so great, however, that they were pressed down upon 
the mortuary bowls, and many of these fragile objects were thus 
broken into fragments. 

Thirty-one skulls in good preservation were taken from the ceme- 
teries at Chaves pass. These are numbered in the National Museum 
catalog from 157669 to 157699, inclusive; but a large number of skulls 
and skeletons were abandoned, as they were in too poor a state of 
preservation for craniometric study. The facial bones of several 
crania from Chaves pass were stained green with carbonate of copper, 
and there were traces of black pigment on others. 

The situation of the ruins at Chaves pass, which are, as has been 
stated, practically on the trail from the Little Colorado valley to the 
Verde, naturally leads to a comparison of the pottery objects from 
the two localities. Fortunately, a considerable collection of Verde 
valley pottery made by Dr Palmer is now in the Smithsonian Institu- 
tion, and affords abundant material for a comparative study. There 
is so marked a similarity between the ancient pottery from Chaves 
pass and that from the Verde ruins, which in turn is related to that 
of the Gila-Salado basin, that it almost amounts to identity. It 
would be impossible from the character and color, as well as from 
the decoration of the mortuary ware from these two regions, to dis- 
tinguish them. The same red ware with rude geometric decorations 
exists in both valleys. There is no doubt that the ancient people of 
the Verde valley were closely related to the builders of Casa Grande 
and the ancient dwellings near Tempe. a 

Ruins Between Winslow and the Hopi Pueblos 

Under this heading are included the remains of habitations on the 
banks of the Little Colorado and its tributaries which were visited on 
trips from Winslow to Tusayan. The author followed the river for 
many miles in order to avoid the Moqui butte, and while he did not 

"The ruins in the upper Gila valley, called Pueblo Viejo, were found in explorations in 1897 to 
be of the same character as those of Chaves pass and Verde valley. 


go as far down as Voltz crossing, he forded the river only a few 
miles above that place. 

The hills hounding the valley retreat a considerable distance from 
the banks of the river in that section of its course, and the road winds 
through a level plain destitute of rocks suitable for building pur- 
poses. At certain points, however, the author passed low mounds, 
not accurately mapped, upon which w r ere scattered fragments of pot- 
tery, most of which was of rough manufacture. These mounds may 
have been sites of small adobe buildings which have weathered away, 
leaving only piles of soil. He attempted no excavations and found 
no standing walls of adobe or stone, but the presence of fragments 
of potter} 7 in quantity would seem to indicate former habitation. 

It would be instructive to dig into one of these mounds, which are 
undoubtedly artificial in character, in order definitely to determine 
their character, which it must be confessed is now highly problematic. 

Although the cavate ruins near Flagstaff and the ruins near the 
Black falls were not carefully examined until 1900, they are described 
here for comparative purposes. 


The following account of these ruins and of those near Black 
falls was published in the American Anthropologist in 1900 (volume 
2, page 423) : 

Sitgreaves, in 1852, seems to have been the first writer to refer to 
the ruins about Flagstaff and along the Little Colorado. He figures 
one of the ruined pueblos near the cascades or falls, a a ruin of the 
same general character as those near Black falls, which he probably 
did not visit. Major Powell, in 1885, visited and later described 6 the 
cliff houses, the cavate rooms of the volcanic cones, and several 
pueblo ruins north and northeast of Flagstaff. He did not visit the 
Black falls ruins, which are undoubtedly similar to some of those 
which he describes. Since Powell's description the literature of 
the Flagstaff ruins has been confined mostly to popular newspaper 
articles, archeologists seeming to have paid little attention to this 

The cavate rooms near Flagstaff are excavated in the lava, or vol- 
canic breccia, and may be classified as (1) cavate rooms with vertical 
entrances, and (2) cavate rooms with lateral entrances. The former 
are well illustrated by the " Old caves," 9 miles east of Flagstaff; the 
latter by the "New caves," 12 miles from the same place, in the same 
direction, and by cavate rooms half a mile west of Turkey tanks. 
These two types of cavate rooms are similar, and their former inhab- 
itants were apparently of the same culture. Major Powell learned 
from the Indians of Cataract canyon that the ancestors of the Hava- 
supais occupied these cavate houses, and lie states that "they doubt-. 

« Probably the "cascades' 1 were the Grand falls, rniles above the Black falls. 
''See Seventh Annual Report of the Bureau of Ethnology, 1891. 


less lived on the north, east, and south of San Francisco mountain 
at the time this country was discovered by the Spaniards, and they 
subsequently left their cliff, and cavate dwellings and moved into 
Cataract canyon, where they uoav live." 

The fragments of pottery seen about the entrances to these caves 
are identical with those found near the pneblo ruins in the neighbor- 
hood, and there is no doubt that the cave inhabitants had burrowed 
in the lava as the most practical means of constructing habitations 
in this neighborhood. Free walls are found in combination with the 
caves, but these walls have no distinctive characteristics, save that 
they are built of lava. This would indicate that the builders simply 
utilized readily available building material * and took advantage of 
peculiar geological conditions. 

Old Caves 

The "Old caves" lie near the top of Old Cave mountain, about 9 
miles northeast of Flagstaff, and cover an area of about 5 acres. On 
the top of this height there is a level space which was surrounded 
by a rough wall made of volcanic breccia, from which a good view 
can be had of the surrounding country. The caves are found on the 
southern slope, and were excavated in a conglomerate of cinders or 
volcanic breccia which bears every evidence of having been erupted 
from a crater or blowhole (plate i). Clambering over the rough lava 
blocks, one finds everywhere on the surface the remains of walls 
indicating former rooms. In places there are level spaces which 
seem to have been plazas, and the entrances into the subterranean 
rooms often open vertically from these levels. At other points it 
would seem as if the walls formed complete rectangles, but there is 
no apparent evidence that they had roofs, which, however, may have 
existed in former times. In fact, the Old caves show combinations of 
underground cavate rooms with free walls above, and when inhabited 
the settlement must have looked like a collection of low one-story 
rooms continuous for several hundred feet. We may therefore call 
this cluster of cavate rooms a pueblo in which each room above 
ground had a corresponding subterranean chamber hewn out of 
volcanic breccia. 

One of the best-preserved and most characteristic rooms of the old 
caves, with a vertical entrance, is shown in the accompanying plan 
(figure 3). It will be seen that there are two subterranean rooms, A 
and B, each of which is entered by an opening in the roof, indicated 
by a dotted line. Room a measures 12 feet each way, and the entrance 
measures about 6 feet. This entrance has a square enlargement, or 
chimney, on one side, which extends to the floor of the room and has 
perpendicular, regular walls. 

At one corner of room A there has been hewn out of the lava a 
small recess (e), the floor of which is lower than that of the room. 
There is also a small recess (f) at one side of the chimney. 




Room B is larger than room A, being about 16 feet square. It com- 
municates with room A by a broken doorway, and has an opening 
through the roof. The floor is somewhat lower than that of A. A 
recess (d) on one side of this room also communicates with the out- 
side by a small opening which bears the same relation to room B that 
the flue F does to room A. 

Room C is an oblong, irregular, subterranean chamber, 5 by 7 feet, 
with passageways into rooms A and B. The opening into a is almost 
perfectly square, that into B less regular. Its floor is several feet 
lower than the floors of the two other large rooms in this cluster. 
There are evidences of clay plastering in several places, and appar- 

Fig. 3. Plan of an " Old cave " dwelling. 

ently the floor, walls, passageways, and possibly the roof, were 
smoothly finished. The plastering has, however, fallen, exposing the 
rough lava corners. 

New Caves 

The mountain in which the New caves occur is about 3 miles west 
of Turkey tanks, or about 12 miles east of Flagstaff. This height is 
interesting from a geological point of view, it being a section of the 
rim of an old crater, as may be seen from its summit. The remain- 
ing portion of the crater rim, that on the eastern side, has been 
eroded into hills, the relation of which to the crater is recognized 
only by their positions. The highest part of the rim, that in which 
the caves are found, is the western wall of the crater, which, with an 
adjacent southern section, forms a crescent, connected by a ridge of 
less altitude. The more northern of these elevations is the higher, 
and the cavate rooms occur on its eastern side. 


From the west the ascent to the mountain, though steep, is not dif- 
ficult, the trail passing stunted cedars growing on a mass of cinders. 
In the depression between the two hills which form the crescent we 
find rows of volcanic breccia fragments arranged in rectangular and 
other forms, suggesting a reservoir. From this point the ascent be- 
comes more difficult, and as one reaches the top of the higher hill he 
finds himself on the rim of a former crater. On the east the rim rises 
almost perpendicularly, and its walls on that side are outcroppings of 
exceedingly rough cinder conglomerate. In this almost perpendicular 
wall, facing what may have been the middle of the former crater, tier 
upon tier of cavate rooms (plates II, in), irregularly arranged and 
very difficult of approach, have been excavated. The crest of this, as 
well as of the adjacent lower section of the crater rim, is capped by 
artificial walls of considerable height, indicating former houses. The 
whole aspect of the place is one of desolation, and the lava appears 
as if it had been molten but a few generations ago. It may have 
been great stress of danger which drove the aborigines to seek homes 
in this forbidding locality. 

Turkey Tank Caves 

About half a mile west of Turkey tanks (about 15 miles east of 
Flagstaff) there is a collection of cavate rooms with lateral entrances 
arranged in tiers. These caves, although not so numerous as the New 
caves, are comparatively well preserved. They are situated a short 
distance to the left of the road from Flagstaff on the uplifted out- 
crop of what appears to be an old volcanic blowhole, and are con- 
fined to the northern side of the depression which marks the former 
place of eruption (see plate iv). 

The outcrop on this side of the depression is composed of alternate 
layers of hard lava and volcanic breccia. The former would tend to 
resist any working with primitive implements, but the latter could 
readily be excavated with stone tools. The average thickness of the 
layers is about 8 feet. Ity the excavation of the breccia the layer of 
harder lava above it has been undermined and has now fallen in 
places, filling the rooms or closing their entrances so that the form 
and dimensions are no longer determinable. As the layers are 
uplifted, vertical entrances into these cavate chambers are absent, 
the doorways entering horizontally from the side of the cliff. There 
are at least three tiers of these rooms, corresponding with the strata 
of volcanic breccia. 

In some of these cavate rooms there is a combination of stone walls 
and excavated chambers, the rooms having been separated laterally 
by a plastered wall of small bowlders brought from the bottom of the 
adjacent depression. Apparently, also, walls formerly existed in 
front of the entrances to the caves, but of these the greater part have 
fallen, and their outlines are difficult to trace except in small sections. 


Entering by a side opening, one passes into a subterranean room 
(plate v) 12 by 10 feet and 6 feet high, the walls and floor of which 
are partly plastered. This room has five smaller rooms leading from 
it, which will be called B, c, D, E, and F. They average about 5 feet 
in diameter, and have their floors depressed about a foot below that 
of the main room, A. The entrances into these lateral rooms, especially 
that into D, are carefully made and almost square, and when plastered, 
as there is good evidence that they once were, made good doorways. 
In fact, although the walls of most of these cavate chambers are now 
very rough, and the rooms seemingly desolate as places of habitation, 
they must once have been comfortable abodes, for the plastering made 
the finish almost as smooth as that of any wall which could be con- 

Several of the rooms in which the plastering still remains have 
ledges and cubby-holes in which the household utensils were doubt- 
less kept (plate vi). The similarity of these cavate chambers to 
those excavated in volcanic tufa in Verde Valley is apparent. The 
material in which they occur is different, but the plans of the rooms 
are almost identical. Whatever peoples inhabited the cavate dwell- 
ings of the cinder cones near Flagstaff and the tufa mesas of the 
Verde, their culture was not radically different. 

Location and Prfvious Exploration 

It has long been known that the banks of the Little Colorado and 
neighboring mesas were sites of ancient dwellings, but exploration 
has been confined mostly to the upper part of the river and its tribu- 
taries. The numerous ruins along the stream from Grand falls to 
its confluence with Rio Colorado have been wholly • neglected, but 
there is little doubt that future excavation will be rewarded with 
many novelties. 

The Black falls ruins have been known for several years to local 
amateur archeologists, and a considerable collection of ancient objects 
has been taken from them by Mr Benjamin Doney, of Flagstaff. 
Under his guidance several well-known residents of that town, among 
whom may be mentioned Dr Robinson and Mr Jack, have visited 
and photographed them. a Herders and cowboys are acquainted with 
the ruins, and the former have cleared some of the rooms for use in 

The geological features of the region in which these three groups of 
rains occur are instructive, but for present purposes one or two simple 
si atements about them will suffice. The two well-marked formations — 
lava and sandstone — have affected the appearance of the ruins. 

a The author was guided to these rains l>y Mr Doney. He is indebted to Dr Rolnnson and 
Mr Reed for kodak photographs, and to Mr Jack for measurements of several rooms. 


The black lava covers the red sandstone, forming great mesas or 
isolated buttes, the summits of which are crowned with ruins. The 
lava ruins have low, rough walls, in which adobe mortar was not 
detected.. The red sand stone "formed a more tractable material, and 
the buildings constructed of it show fine masonry with adobe mortar. 
These ruins ordinarily stand on the brinks of small canyons eroded in 
the sandstone, on isolated blocks of the same stone, or on ridges left 
by erosion. If these lava and red sandstone ruins were found in dif- 
ferent localities they might be regarded as products of different peo- 
ples, but their existence side by side in this region shows that the 
slight differences in their architecture were due simply to the build- 
ing materials employed. The irregular forms of the lava blocks made 
it impossible to construct from them the fine rectilinear walls which 
were possible with the well-squared blocks of sandstone. The erosion 
of the lava produces a coal-black, porous sand, which as a rule covers 
the finer red soil derived from the sandstone. This soil, drifting into 
pockets or depressions in the surface rocks, afforded burial places for 
the inhabitants of the villages. 

This region has few trees; there are no pines, and only a few cedars. 
It is the same sagebrush country which we find near the upper Little 
Colorado at Holbrook and Winslow. 

The region is arid; it now has few springs, those which were 
used in ancient times having probably been filled by drifting sand. a 
Volcanic agencies have left their mark on the whole region, causing 
in places deep fissures in the rocks, into some of which a strong cur- 
rent of air continually passes, and from one of which emerges a roar 
as though of subterranean currents of water. One of the largest of 
these fissures is about 2 miles from the Tuba road, on the way to 
the ruins called group A; others are found in the rocks near ruins 
G and H of this cluster, where their depth has not been determined. 
These crevasses, which are no unusual feature in the geology of this 
region, vary in breadth from a few inches to many hundred feet, and 
from a hundred yards to miles in length. When very broad they 
form canyons which end abruptly or merge into "washes" as the con- 
figuration of the country may dictate. 

General Features 

The ruins near Black falls are as a rule rectangular in form, with 
similarly shaped rooms of one or more stories. Curved walls are 
rare, although in some instances the shape of the ruin follows the 
curvature of the mesa on which it stands. As has been stated, the 
ruins are built of both sandstone and lava, and the two varieties are 
found in close proximity, sometimes within a few hundred feet of 

« The author does not share a common belief that when these now ruined structures were 
inhabited the precipitation was greater. In an arid region springs are rapidly filled by drifting 
sand if not dug out repeatedly. The Hopis are obliged to clean out some of their largest springs 


each other. The character of the sandstone of the region is such 
that when the stone is fractured slabs are produced which make pos- 
sible the construction of excellent walls. Blocks of lava, however, 
have no flat faces, and their use as building material results in poor 
masonry, for the adobe mortar readily washes from the joints and the 
walls soon fall. It is rare to find houses built of lava which now 
stand many stories high. The best rooms constructed of lava con- 
tain also sandstone slabs, which have strengthened their walls, as in 
the "Citadel" of the Black falls ruins, where blocks of sandstone 
were also used as lintels. None of the walls show evidence that the 
building stones were dressed after being quarried. 

The sites of these ruins areordinarilv elevated, and it is not uncom- 
mon to find an entire mesa top either covered with rooms or sur- 
rounded bv a wall. 


The highest walls of these pueblos were as a rule situated on the 
north and west sides, the pueblos being terraced on the south and 
east. This arrangement was apparently adopted to secure sunny 

The ground-floor rooms had no lateral external entrances, but where 
there were several chambers side by side they communicated with each 
other by doorways. In the case of two or three story houses, it is prob- 
able that the ground floor was used for storage and was entered from the 
roof. This is an architectural feature still retained in the old Hopi 
houses, but it has been somewhat masked by modern buildings erected 
in front of them. The old houses of Walpi, Sichumovi, and llano 
had ground floors Avhich were entered from the roofs, to which one 
mounted by ladders, while entrance to the second story was gained 
by means of a side doorway from the roof of the first. Many of these 
old rooms are still to be seen at Walpi, especially around the plaza, 
and there are one or two examples in the villages of Sichumovi and 
llano." The oldest houses of Tusayan never had lateral entrances 
from the ground floor, but when the first story was occupied it was 
provided with a hatchway in the roof. This type of room, however, 
i> rapidly disappearing, the majority of ground-floor rooms on the 
East mesa now being provided with doorways in the walls. On the 
Middle mesa and at Oraibi the number of ground-floor rooms entered 
by a side door is still smaller than on the East mesa. It may safely 
be laid down as a rule that whenever in the Hopi pueblos one finds 
rooms on the ground floor entered by lateral doors, the construction 
is new.* 

" A good example of the ancient houses of Walpi. in which the lower story serves as a dwelling 
room at certain times, is Saliko's home, near the Snake rock, and the row of rooms from Honsi's 
house to the Mon kiva. The Flute house is also a fine example of this type. In Sichumovi the 
house of Ptltce illustrates this ancient type, and there are several examples of it in Hano, of 
which Kalacai's house is a good one. 

b The author will consider this architectural likeness of the ground rooms of the ancient ruins to 
old Hopi houses in his final article on the Black falls ruins, where plans will he given illustrating 
the relation of the ground-floor rooms with lateral doorstothe old rooms on the East mesa. The 
ruins near Black falls have their ground-floor rooms like the old rooms of the Hopi pueblos. 


Iii many of the ruins there are found at the base of the mesa on the 
south and east sides rooms of a single story which, from their position, 
we may designate basal rooms. The}^ are now covered with debris, but 
were once protected by the overhanging edge of the mesa, suggesting 
cliff houses, of which they may be a survival. These basal structures 
may have been used as granaries, but in none of them were remains 
of roofs found. 

With the exception of ruin A, group B, most of the ruins show little 
evidence of long occupancy; few logs or beams remain in them, there 
are no extensive deposits of debris, and there is a lack of large quan- 
tities of pottery fragments such as are usually found about pueblos 
which have been occupied for many generations. The general indi- 
cation is that these buildings were inhabited in comparatively modern 

None of the rooms show marks of surface plastering, except those 
of group B, where it is confined to the interior of the walls, as is the 
case with the older Hopi buildings. 

The size of the rooms is much greater than is. common in very 
ancient ruins. No kivas are found, and it is believed that the 
religious ceremonies were held in the ordinary domiciles. No build- 
ing had a roof intact, but in many instances tlie remains of the roofs 
and floors of the upper rooms were found in the chambers below. 

The fact that wooden beams occur so abundantly in ruin A, group 
B, implies that it was either the last pueblo to be abandoned in this 
neighborhood or that the beams were taken from the others to it, and 
when it was deserted its inhabitants moved too far away to carry heavy 
objects with them. Some of the timbers in the modern Hopi houses 
are,said to have been dragged from the Little Colorado, possibly from 

old ruins. 

Group A 

Group A includes a cluster of ruins which as a rule are small and 
have a general similarity in construction. It is situated about 15 miles 
west of Little Colorado river. Following the road from Flagstaff to 
Tuba to within about 11 miles of Tanner's crossing, after passing 
Deadinan's flat the visitor turns to the right, and, proceeding 4 miles 
eastward, finds himself in the midst of the group. There are no trails 
or wagon tracks from the well-traveled Tuba road to group A, but 
the country is so level that one can readily go overland to almost 
any point. A castellated, truncated lava cone, the " Citadel r of 
the group, can be seen soon after one leaves the Tuba road, and 
this prominent landmark gives the general direction of the ruins 
among which it is situated. From the top of this citadel all the ruins 
of group A, with one or two exceptions, are visible, and the visitor is 
advised to inspect it first in order to determine the position of the 
surrounding ruins (see plate vn). 





The walls of the Citadel (figure 4) are constructed of blocks of lava 
and sandstone, and cover the top of a truncated elevation. They are 
arranged about a level central court or plaza, the surrounding walls 
of which are best preserved on the western side. The hill on which the 
citadel is built bears evidence of having once been a volcanic cone, 
and Avas an advantageous place of refuge for the inhabitants of the 
neighboring houses, as it had a commanding position, was difficult of 
access, and was well fortified. As some of the structures were of two 
stories, they appear to have been permanently inhabited. 





I |H 

Fig. 4. The Citadel, group A. 

Twenty-three small ruins were counted from this elevated position 
(see plate vn). For convenience of description these may be desig- 
nated a, b, c, etc. 


Ruin a of group A is situated at the base of the truncated mesa of 
the Citadel. It is built of red sandstone, with a few courses of lava 
blocks, is 50 feet long by 12 feel wide, and contains five rooms arranged 
side by side. Although the house was evidently nevermore than one 
story high, the many fallen building stones would seem to indicate 
that its walls were once considerably higher than al present. Few 
floor beams or rafters were detected. 


Near this ruin, at the base of the hill, are four walled inclosures, 
one above another, suggesting terraced gardens. Their low walls are 
composed of alternate rows of lava and sandstone. Near these former 
gardens is a depression which may once have been a reservoir. This 
ruin is the onlv one visited which was not built on an elevated mesa 
at or near the edge of a canyon. 


There are remains of three houses, built of lava and sandstone 
blocks, on a small lava hill a few hundred feet north of the Citadel. 
On the same elevation there is a circular wall which may have served 
as a fortification. Most of the walls of the ruins have fallen, and it 
is almost impossible to determine the relationship of the former 
rooms. There are also some small ruins on a lava hill near the 
elevation on which B, c, and D are situated. 


A considerable distance from the last-mentioned cluster, but in the 
same direction from the Citadel, there are situated two conspicuous 
ruins visible from a considerable distance. One* of these, on the top 
of a lava mesa, is built of the same material of which the mesa is 
composed; the other, situated at its base, is constructed of red sand- 
stone. Near the latter, on a lava mesa, there are many pictographs, 
representing spirals, frogs, snakes, and unknown figures. There is 
much broken pottery near ruin F. 


These ruins, especially G, H, and J (plates viii-xi), are constructed 
of limestone, and are situated on the brink of a canyon, at the bottom 
of which, near ruin G, are mounds indicating the site of I. The 
walls of G, H, and J are well preserved, and show some of the best 
aboriginal masonry in Arizona. 

Ruin G (plate viii) had two rooms with walls rising 20 feet from the 
rim of the canyon. The lower courses of the walls are much larger 
than the upper, as is true of others in this neighborhood. The level 
of the floors is indicated by courses of larger stones. 

Ruin J (plates x, XI, figure 5) is the best preserved of all the ruins 
in group A, and presents exceptional features. It is situated on the 
left wall of a canyon which is about 40 feet deep and equally wide. 
It deepens and widens east of the ruin, and then narrows, forming a 
natural corral inclosed by cliffs. Eight good rooms were noted in that 
part of the ruin situated on the top of the canyon wall, and in the 
canyon below it there were several semicircular basal rooms, some of 
which were sheltered by an overhanging cliff. Similarly sheltered 
rooms are found in many of the ruins in this neighborhood, but 
nowhere else are they so well preserved. There are no beams in 




place, but their former positions are shown in many walls by 
openings, indicating that when inhabited the pueblo had two, 
possibly three, stories. An inclosure which may have been a ninth 
room is so filled with fallen walls that the details of its construction 
or size could not be determined. 

As none of the rooms have external lateral openings on a level with 
the foundations, it is naturally supposed that all were entered by 
means of ladders and hatchways. There are a modern doorway and 
fireplace in one room, evidently of later construction than the walls. 

Perhaps the most problematic structures in this ruin are the small 
cvsts in the canyon walls east of the entrance. A thin layer of 
softer rock has so weathered as to leave a horizontal crevice which at 
intervals is divided by stones set on edge into receptacles a foot or 
so deep. The}' were formerly closed by flat slabs of stone, only two 
of which now remain in place. These cysts were nicely plastered, 
and the slabs which closed them were luted in place with adobe. 
Nothing was found in them to indicate their use, whether as burial 


basal Roonr 

^ . M « ^|\l»"' ' 

. \ v ^*wiA...^>.;.iM\iG>iii"'>- ,,1, ' l '' ,v! "'""' "'"• '■" ■?;^^\ 



i t 

fir vy.'tft '/ 

I • p ■ " ■ 




iff eo'h itf/'jW * g m 



IOF 1 


Fig. 5. Plan of ruin j, group a. 

places for the dead or as bins for the storage of corn. Their number 
was considerable, but they were so small that their capacity could 
scarcely have been more than a few bushels. This is the only ruin in 
which such inclosures were found, and no theory is advanced as to 
their former use 


Ruin K, which evidently formerly contained several rooms, is 
divided into two sections and is situated on a high lava mesa difficult of 
approach. The walls of the larger section inclose three well-preserved 
rooms, and still rise to a height of about 8 feet. Five feet above the 
base the red sandstone blocks of which the walls are built are 
replaced by a course of stone of lighter color, which forms a horizon- 
tal band around the ruin. The second section consists of a low, 
rough wall built along the edge of the cliff, inclosing a level space in 



[ETH. ANN, 22 

front of the first section. There are isolated rooms in this inclosure, 
and a depression which may have been a reservoir. This ruin, like 
many others, consisted of dwellings and a fort for protection. There 
are instructive pictographs on the rocks near by. 


At the base of the mesa on which the last- mentioned ruin stands 
is a ruin of red sandstone with five rooms and a foundation of 
unusual shape. A huge rock, cubical in form, has fallen a few yards 
from its former position in the bluff. Ruin L is built on the top of this 
detached block, and its fairly well preserved walls are separated 

Fig. 6. Section a, ruin A, group b. 

from the bluff on all sides by a wide crevice. From a distance the 
ruin appears to be perched on the bluff, but closer observation shows 
its separation from the latter by an impassable natural moat. 


This is an oblong ruin rising from the side of a deep, narrow 
canyon, with walls consisting of alternating courses of large and 
small blocks of red sandstone. Some of the walls have fallen, but 
sections fully 10 feet high still remain in place. There are evidences 
of five rooms, each two stories high, but most of the chambers are 




filled with fallen stones. The cemetery of this pueblo lies west of 
the ruin, where there are also remains of walls. 

Small ruins may be seen near the road from group A to group B, 
a few miles to the left. Their walls are in good condition, but no 
peculiar features were observed. 

Group B 

ruin A 

The largest of all the ruins in the Black falls cluster, and one which 
bears evidence of having been inhabited for a considerable time, lies 
about 35 miles northeast of Flagstaff and about 8 miles from the 
Little Colorado. This structure is built on a ridge of sandstone 
extending in a northeast-southwest direction, and consists of two 
large buildings of moderate elevation (plates xii-xvi, figure 6). On 
each side the ridge slopes gradually to a depression, the talus on the 

^ul'i.Ui*" 1 ''""' 

■■j£" •-,■"</"■.„., 

""■/TV . 

~ * - '■•' 


£%. ...-•■ 


i * 


nriv • .• 

. „.»«*\l!|>u..... 


_• v - --> 

Fig. 7. Plan of section A, ruin A, group B. 

east covering a series of rooms, while on the west side, where the 
slope is more abrupt, no rooms were discovered. The rnin is divided 
into two sections connected by rows of one-story rooms, the walls of 
which have fallen. Remains of a great number of roof and floor 
beams are still scattered throughout the debris. These beams are 
larger than those in any other rnin of the same size known to the 

It is difficult to determine the original number of rooms in the first 
section of this ruin, as the tops of the walls have fallen, filling the 
chambers with debris. How many basal rooms were buried in the 
talus of fallen walls at the base of the mesa on the eastern side could 
not be discovered. Room a of this section (see figure 7) is elevated 
on a rocky base about 10 feet high. The chamber is small, and its 
walls have fallen on two sides. The debris has been cleared out of 
this room by Mr Doney, who found in it the desiccated remains of 
an infant wrapped in four well-preserved cotton blankets. 



[ETH. ANN. 22 

Room B is a small, narrow chamber with good walls on three sides, 
but the fourth wall, which was situated on the edge of the mesa, has 
fallen over the brink. 

The ground-floor chamber of room c is formed by a gap in the mesa, 
from which a large cubical block has fallen. The walls of this cham- 
ber are the natural rock, to the surface of which adhere fragments of 
plastering. The beams of the floor of an upper room still rest on the 
edge of the gap, as in some of the kivas of Walpi to-day, especially 
those on the eastern edge of the mesa. These are built in a depres- 
sion, the solid rock forming the walls on three sides, the fourth wall 
being of masonry. 

Room D is buried under debris, and the broken beams, which have 
pressed down on a plastered banquette, are still visible. The reeds, 
straw, and impressed clay which once formed a floor may be seen in 


'''/•/mijm-.i h , ||, ,> l ' '''•/. /,<! i'>i..,'/„ „.,,..„., 




< ■--<- lfe><:'. 

"'iWwMi/jlUov.mirf,, „ 

"""•^>"" , '"""^| l |. 1 ^nV rt> ^"»m 11 , IM|ln , ii| 

Fig. 8. Plan of section B, ruin A, group B. 

Room E has two stories, and the floor beams and rafters are still in 
place but buried under debris. A high wall extends from the east- 
ern wall of room E, crossing a depression in the cliff, which is. bridged 
by logs serving as its foundation. 

It seems within the bounds of probability that there were 30 rooms 
in the first section of ruin A, group B, including the basal rooms now 
deep beneath the fallen walls of the higher portion of the ruin. On 
the supposition that half of these were uninhabited, and that there 
were four persons to each room in the remainder, the first section of 
the ruin would have housed a population of 60. This, however, on 
the basis of the present population of Walpi, as compared with the 
number of rooms in the pueblo, is a rather low estimate. Considering 
the population of the second section as about the same as that of the 
first section, and that of the connecting rooms as about 30, the approxi- 
mate population of the pueblos would have been 150. Estimated on 
the basis of that of Walpi, the population would be 200. 


The rooms of the second section (figure 8), several of which are well 
preserved, are lower than those of the first section, and the detritus has 
covered the base so completely that the mesa is inconspicuous. Room 
a (plate xiv) is nearly square and is built on two rectangular rocks, the 
top of which forms the floor. One of these rocks forms a side of the 
lower story of the adjoining room B, which is in the best condition of 
any in this section. The walls of this room are well preserved, and 
it was occupied as a habitation by a herder a few winters ago. There 
is a lateral doorway through the wall on one side, and in one corner 
is a fireplace communicating with a chimney, which will later be 
described. This room is 12 feet 4 inches long by 9 feet 7 inches wide. 
In the second section many walls are still standing high above their 
foundations, indicating rooms now filled with fallen debris, in which 
beams, fragments of pottery, and other objects may be seen. Ten 
large rooms were counted, several of which had two stories. As has 
been stated, there were apparently basal rooms on the eastern side. 
The entire section is about 60 feet long. 

A chimney-like structure (plate xv) is one of the most conspicuous 
objects in this part of the ruin. It rises from the mass of debris 
covering room E and communicates with the fireplace in room B, but 
a vertical line from its top is 7 feet 10 inches from the nearest wall of 
the room in which the fireplace is situated. Whether this chimney is 
aboriginal or not, or whether it is a chimney at all, are open questions. 
Excepting its state of preservation and fine masonry, no evidence was 
found that it is of more recent date than the walls of the rooms. If 
it is an aboriginal chimney, which is doubtful, its structure is unique. 
It may be a ventilator, comparable with the chimney-like structures 
described by Mindeleff in the kivas of Canyon de Chelly. 

One of the finest reservoirs (plate xvi a) which the author has seen 
in connection with a ruin was discovered near the bottom of the ele- 
vation on which ruin A of group B is situated. This reservoir is cir- 
cular in shape, 50 feet in diameter, and carefully walled. It lies south 
of the second section of the group, and apparently had a break in the 
wall in line with the depression east of the ruin. It appeal's to belong 
to 1 lie same type as those reservoirs on the East mesa of the Hopis in 
which snow and rain are collected for future use. 

There are instructive petroglyphs near ruin a, group B. A number 
of rock etchings observed in a small canyon about a mile from the 
ruin werei>ecked in a perpendicular wall, protected by the overhanging 
rim of the canyon. These petroglyphs were evidently made by the 
former inhabitants of this region, as one of the best examples shows 
the same design as that figured on pottery from the neighboring ruin. 
There were Likewise butterfly, sheep or antelope, and other figures. 

It would be quite impossible in this preliminary notice to give a 
complete account of the archeologic objects which Mr Doney has 
taken from this ruin, but even a preliminary sketch would be incom- 

22 eth— 04 4 


plete without some reference to them. One of the most important 
objects is the desiccated body of an infant wrapped in coarse cotton 
cloth,. allusion to which has already been made. This bundle was 
inclosed in three small cotton kilts which were later washed and 
found to be "as good as new." At the foot of the infant was a desic- 
cated parrot (?), some of the brilliant plumage of which is still to be 
seen. This bird has a prayer stick tied to one leg, which makes 
reasonable the belief that it w T as a ceremonial object. Another inter- 
esting specimen in the Doney collection is the dried body of a dog, 
which was found in one of the deep clefts in the rock near one of the 
ruins. This dog has a head similar to that found by the writer in 
the Chaves pass ruin. There are also several fragments of beauti- 
ful cotton cloth and netting. Some of the specimens are embroidered, 
others are painted with circles and other geometric designs. A heavy 
wooden club, several planting sticks, and other wooden objects are to 
be seen in Mr Doney 's collection. There are also many cigarette 
canes, some with woven handles, as well as seeds of cotton, squash, 
gourd, and corn, and many objects of shell, as tinklers, ornaments, 
rings, and bracelets. One of the best Haliotis shells the author has 
ever seen from a ruin was found in one of the graves. 

There are also many large turquoise ornaments, some an inch or an 
inch and a half square. The many metates are made of lava, and 
are deeply worn, as if from long use. A copper bell from a grave 
near ruin A is a remarkable specimen. It has the same form as the 
bells from Arizona ruins, which the author has elsewhere described, but 
on one side are ridges indicating eyes, nose, and mouth, apparently 
made of strips of metal soldered or brazed to the surface. It is not 
believed that this bell was the product of the former occupants of 
these now ruined structures; more probably it was obtained by them 
through barter. 


Across the depression north of ruin A, beyond the reservoir and on 
top of a mesa, there is a rectangular ruin consisting of two sections 
connected by low, parallel walls, which inclose a rectangular plaza. 
It appears that each section was composed of two single-story rooms. 
No beams or other evidences of roofing are now visible, but a consider- 
able quantity of masonry has fallen into the inclosures. From the 
base of the mesa to the ruin an old trail can be traced by rows of 
stones on the eastern side, and on the same side there are likewise 
remnants of rooms. Graves were found among the rocks at the base 
of the mesa. 

ruin c 

About half a mile north of ruin A of group B there is a fortified mesa 
with several rooms, some of which had two stories. The surface of 
this mesa is flat, the rim is round, the sides are perpendicular, but 




of moderate elevation. Most of the walls built on the rim, continuous 
with the mesa sides, have fallen, but sections of the houses 10 feet 
high still remain, and the roof beams and wattling may be seen in 
place in one or two rooms. 

There are some fragments of broken metates made of lava, many 
potsherds, and a considerable pile of debris at the base of the mesa. 
Ruin A can be seen from the highest point, and the distant ruin A of 
group c is plainly visible. The cemetery is on the east side, among 
the rocks at the base of the mesa. 

Fig. 9, Ruin a, group c, from the south. 

Group C 

ruin A 

This ruin, which lies 40 miles by road from Flagstaff and 5 miles 
due west of I Hack falls, is one of the most impressive masses of abo- 
riginal masonry in this section (plates XVII-XIX, figure 0). It is 
visible for many miles, and from a distance resembles an old castle 
as it looms from the north end of an isolated, oblong, red-sandstone 
mesa rising 15 feet above the plain. The south end of the mesa is 
higher than the north extremity, and on its rim appears to have been 
built a low wall inclosing a plaza. Standing walls cover about one- 
half the surface of the mesa. On its east side, about midway of its 



[ETH. ANN. 22 

length, there is a gap with perpendicular walls extending about 14 
feet into the side and almost bisecting it. 

The following measurements of ruin A, group c, were made by Mr 
Jack, who has kindly placed them at the author's disposal: 

The longer axis of the mesa bears north 10° east (the bearing was 
obtained by using the face of the eastern wall of the highest building). 
The width of the mesa at the middle point, measured from the rim of 
the overhanging cliffs, is about Go feet. The height of the tallest 
wall of room A is 19 feet above its foundation, on top of the mesa, 
which is about 15 feet high. The inside measurements of the same 
room are: Top of mesa to probable position of first floor, 7 feet G 
inches; first floor to probable position of second floor, 8 feet; bench 
on which the floor beams of the second floor rest to the top of the 

t 0"> l " t(| '"''"'// 








,U> o * - "-Cu 

r ^ o 


v «1. 


,0 c> 

^ 7 

3 ^vXJj 





rc -^ r ■ 





^ « 

ess**. HOQMS 
Fig. 10. Plan of ruin A, group c. 

wall, 3 feet. It may reasonably be concluded that the third story 
was as high as either of the other two, or about 7 feet 6 inches, which 
would make the original height of the wall about 23 feet. 

The inside horizontal measurements of the north and south walls 
of room A are not the same. The former is 11 feet 4 inches, the latter 
9 feet 9 inches. The east and west walls are 12 feet long. Room c 
is 17 feet 9 inches long by 9 feet 7 inches wide. 

.Although the standing walls of this ruin are the best preserved of 
any of those examined, no wooden beams were found in place, nor 
were there remants of the flooring or other debris in the rooms them- 
selves. This absence is explained by the supposition that at the time 
of the abandonment of the settlement, or later, the woodwork was 
carried away for use in new habitations. Possibly they were taken to 


ruin A of group B. There is good evidence that this ruin once had 
large floor beams, as is indicated by openings in the walls in which 
thev rested. 

Examination of the ground plan (figure 10) shows that the whole 
surface of the mesa was once covered with rooms, the walls of which 
still extend to its edge. The highest walls, those which surrounded 
room a, are three stories. The two outside Avails rise directly from 
the edge of the perpendicular cliff. There are several small openings 
at various levels, and holes in which rested the great beams that once 
supported the flooring are readily seen. At the corners of the rooms 
the masonry of the second story is bonded to that of the first and 
third, imparting solidity and strength to the high walls. There is no 
entrance or passageway between rooms a and c, but access was had 
to room a from room B. Room B is almost perfectly inclosed by 
standing walls, formerly two stories high. The wall on the north side 
has been overturned, and the many stones which have fallen at the 
base make an entrance at this point possible. As is shown by the 
depressions in the walls, this structure once had two large beams in 
the roof of the first story, but they have disappeared. Room c has 
one story; its walls are complete on all sides, and there is an interior 
entrance into room D, and an exterior passageway. Rooms A, B, and C 
are conspicuous from a distance and form the greater part of the ruin. 
At intervals on the rim of the mesa other walls are found, some sec- 
tions of which are 4 or 5 feet high. It is difficult to trace the walls 
of the rooms designated D and E. This ruin also has cave rooms at 
the south base of the mesa, which recall those of the other ruins in 
the Black falls cluster. 

Plate xvn shows ruin A, group c, from the east. The tall, square 
tower on the left of the plate incloses room a, and the lower wall 
extending to the gap is the side of room c. The fragments of masonry 
on the right of the gap are all that remain of the walls of room E. 
The mounds on the mesa to the right of the last are remnants of an 
encircling wall and of rooms which once surrounded the open space 
on the end of the mesa. On this side of the mesa the upper part 
overhangs the lower, forming a cave, but no indication of rooms was 
detected here. 

The wall on the edge of the mesa which shows at the left of plate 
xviii is a part of room D, and at the bottom of the cliff at this point 
can be seen the walls of the basal rooms built at right angles to the 
cliffs. These are also shown in plate XIX. 

The cemetery is about 100 yards east of the ruin and is small in 
extent. The mortuary objects found in a single grave opened will 
give an idea of the burial deposits. The graves are oval, and consist, 
of cysts made of slabs of stone set on end and covered with other flat 
stones. The upright stones were cemented together with adobe, the 
covering slabs being apparently luted to the edges of the uprights. 
These burial cysts were commodious, and in the one uncovered the 



[ETH. ANN. 22 

body, which was that of a woman, lay on one side, at full length, with 
the head at the wider end. To the right of the hips were found a 
decorated food bowl in which was a smaller bowl, a large and beauti- 
fully decorated vase, and a second small food bowl. On the left arm 
was an armlet made of a Pectunculus shell identical with those found 
in the ruins of Homolobi. On the breast there was a remnant of a 
wooden prayer stick painted green. Near the mastoid processes 
were square ear pendants made of lignite covered with a turquoise 
mosaic surrounding a central red stone. These are beautiful speci- 
mens of turquoise mosaic, far superior to those now in use in the 
Hopi pueblos. The skeleton was in a very poor state of preservation, 
probably because of the character of the soil, which is a cinder sand 
through which water readily percolates. There is a general similarity 
in the texture and decoration of the four pieces of pottery found in 
this grave. They belong to the black-and-white variety and have 
geometrical ornamentation. 

,„,.„,. iiw./aiii nil lui 

1| , llll || ) U/.lW«"" , " , " , " / """"""'% 

. -c v f y • <>■- 1> o c*; 


GASAi, ffQQ/U 



Fig. 11. Plan of ruin b, group c. 


About 2 miles from the large ruin just described, to the left of 
the road to Schultze's spring, is a small red-sandstone ruin standing 
on an isolated bluff. This ruin covers the top of the mesa, and is 
conspicuous for some distance. The rim of the mesa overhangs in 
places, as the lower strata are much eroded, and the ruin can be 
entered at only one point. All the rooms of this ruin are single 
storied, and most of the walls are high, though there is a consider- 
able quantity of fallen stone in the rooms and at the base of the 

Room A (see figure 11) is a semicircular inclosure most of the walls 
of which have fallen. It is perched over a projecting table or plat- 
form, the rim of which the wall covers. The ground plan of room B 
is nearly square ; the walls are well preserved and rise directly from 
the edge of the mesa, which is steep on three sides. The interval 
between rooms B and D is strewn with stones, but traces of low walls 
can be seen. One of these walls is on the edge of the steep mesa; the 


other, parallel with it, almost divides the space in halves. This is the 
part of the ruin which one enters first after climbing up the talus of 
fallen rocks. Room D is large, with well-preserved walls 4 or 5 feet 
high, and with a projecting platform on one side, on which only 
obscure indications of artificial structures may be detected. 

Room F is rather small, with walls built over a projecting platform, 
resembling from below a bow window. Room E is well constructed; 
it contains considerable debris, and its sides are continuous with the 
perpendicular wall of the mesa. At the base of the cliff, just below 
room E, there is a low, almost circular wall, forming an inclosure 
somewhat similar to the basal rooms of some of the ruins already 
described. Although in general its architecture does not differ from 
that of many other rectangular ruins previously discussed, the over- 
hanging platform gives a unique appearance to the structure. About 
300 feet eastward were noted the edges of flat stones which indicate 
burial cysts. The whole length of this ruin is 46 feet, and the width, 
including the projections at f and G, 21 feet. The sizes of different 
rooms measured were : 

Room b, 10 feet 8 inches by 9 feet 10 inches. 
Room d, 15 feet 4 inches by 10 feet 5 inches. 
Room e, 10 feet by 5 feet 10 inches. 

The following bearings were taken from this ruin : 

Group c, ruin a, bears north 12° east. 
Mount Agassiz bears south 48° west. 
Schiiltze's spring bears south 50° west. 


The Hopi Indian Honani declared that there was a large ruin on 
the mesa not far from his house at Burro spring. This ruin, how- 
ever, was not visited, as Honani was away when the author passed 
through that country. There is a legend that some of the clans of 
Shumopovi once lived at this point, which is apparently on the line 
of migration from the ruins on the Little Colorado to the Middle 
mesa of Tusayan. a Honani is a prominent man in Shumopovi, which 
fact may account for his occupation of land near the Burro spring. 

The preceding description will give a general idea of the ruins in 
this section. It is not possible to compare them with the ruins of 
Homolobi, where most of the walls ha ve disappeared or have so fallen 
as to render the original plan unrecognizable. The difference in 
building material employed in the construction of the pueblos on 
Chevlon fork of Little Colorado river must have imparted a some- 
what different character to the buildings erected there, but there is 

" The simple existence of a permanent spring of potable water in this part of Arizona may be 
taken as indicative of ruins in its immediate vicinity, and when such a spring lies on or near an 
old trail of migration, evidence of former settlements can not be difficult to find. Both Big 
Burro and Little Burro springs lie on the pathway of migrations of the southern Hopi clans in 
their journey from Homolobi, and were halting places for longer or shorter periods. 


some likeness between the ruins at Chaves pass and the lava ruins 
near Black falls. In this connection it ma} 7 be stated that there is 
also a large rain near Homolobi built of lava blocks on a lava mesa. 

The racial and clan kinship of the former inhabitants of these pueb- 
los is somewhat problematic, but it is quite likely that the people 
were akin to the Ilopis. This is shown not only by the character of the 
houses, but also by the pottery and various other objects found near 
them. Both legendary and archeological evidences point to the con- 
clusion that the people who once inhabited the pueblos near Black 
falls came from the north, and were related to those who once lived in 
cliff houses and other habitations on the Rio Colorado and its tribu- 
tary, the San Juan. Hopi legends say that the Snake clans formerly 
lived at Tokonabi, on the Rio Colorado, and that they migrated south- 
ward and built a pueblo about 50 miles west of the present Hopi 
towns, which they called Wukoki. This pueblo, it is said, still has 
high-standing walls. The direction and distance of the Black falls 
ruins from Walpi correspond pretty closely with those given in the 
legend, and while it may not be possible to identify any single ruin of 
this cluster as Wukoki, the traditional Wukoki of Hopi legend is not 
far from Black falls. The tradition that the inhabitants of these 
ruins came from the north is supported by the close resemblance in 
character and decoration between their pottery and that of the San 
Juan ruins. 

It might naturally be supposed that there would be a close likeness 
between the pottery of the Black falls ruins and that of Homolobi, 
and that kinship once existed between the inhabitants of these 
pueblos on the same river. Close study, however, shows marked 
differences, and the author is led to the belief that while both were 
pueblo people, and, therefore, similar in culture, the clans which 
inhabited Homolobi were not the same as those which lived in the Black 
falls villages. The clans which lived at Homolobi came from the far 
south, through Chaves pass, while those at Wukoki came from the 
opposite direction. Both evidently sought refuge in the Hopi pueblos, 
where their descendants now live together. The clans from Homolobi 
were the Patki, Patun, and Tabo (Piba), whose route to the Hopi 
towns was by a trail which extends directly north past the " Giant's 
chair." The clans from Wukoki were the Tcua and others who 
migrated almost eastward when they sought their home in Tusayan. 


general features 

The mortuary pottery from the three ruins, Homolobi, Chevlon, 
and Chaves pass, is distinctive and typical, with general resem- 
blances to that from other localities. As a rule it is more varied in 
character than that from the true Tusayan ruins, Shumopovi, Awatobi, 


and Sikyatki, though its decoration has many likenesses to that on 
the pottery from these ruins. It has seemed best to discuss the 
ceramic ware of these three localities together, but in so doing it is 
thought necessary to mention the particular place from which each 
specimen was obtained. a 

It has been shown in an account of the pottery of Sikyatki, where 
conclusions were drawn from a large collection, that there was not a 
single piece of glazed pottery found in that ruin. At Awatobi few 
such fragments were found, but in the Homolobi and Chevlon ruins 
there were many glazed bowls, pots, and jars. 

The question whether the ancient Pueblos glazed their ware has 
been answered in both the affirmative and negative, and this differ- 
ence, no doubt, is due to the want of a good definition of the term 
glaze. Some of the bowls found at Homolobi and Chevlon have a 
black vitreous covering resisting a knife point, but which is not the 
gloss derived from polishing the vessel, but apparently from some 
salt used in the preparation of the black pigment with which the 
ware is painted. This glaze, however, has not been detected on any 
colors but black and green, or on any ware except the red, which is 
. so abundant in both the ruins here described. 

It is hardly necessary to consider at any great length the various 
forms of ancient pottery obtained in 1896, for this would simply 
duplicate work already published in the author's account of Sikyatki. 
Moreover, the question of variety of forms has already been amply 
discussed by others. The mode of manufacture, technic, coloration, 
and like questions were the first to attract attention of students, and, 
while by no means exhaustively presented, are treated more exten- 
sively than the character and meaning of the decoration. A few 
types present the various forms of pottery from the ancient ruins, 
and for a study of form alone the material in our museums is ample. 
With derivation of symbols, however, the problem is very different, 
for in a collection of thousands of specimens we rarely find two in 
which the ornamentation is the same. In a general way it may be 
said that certain decorative types are followed, but the variations are 
so many that in attempting to present an adequate idea of ancient 
ceramic ornamentation it is necessary to describe almost every speci- 
men. Manifestly that would be impossible, and as we need classifica- 
tion in this department of study, the following is proposed. 

« Unfortunately for close sttidy of the lesson taught by Pueblo pottery regarding the migra- 
tion of tin- ancient people of Arizona, the ruin from which ancient Tusayan ware was collected 
is not mentioned in curly writings on old Pueblo pottery. Thus, we find specimens from 
Awatobi, Canyon de Chelly, and Sikyatki given one locality, '"Tusayan, " and modern Tanoan 
pottery made at Hano by colonists from the Rio Grande bearing the same indefinite description. 
Almost all the modern pottery from " Tusayan v in the National Museum is intrusive in that 
province, and is practically modified Tanoan. 



The classification of pottery objects by color and surface finish 
leads us to refer them to the following groups: 1, Coarse unpolished 
ware, undecorated; 2, coarse unpolished ware, decorated; 3, polished 
ware, undecorated; and polished ware, decorated, which maybe again 
divided into: 4, red and brown ware; 5, yellow ware; 6, black ware; 7, 
black and white ware; 8, red and black ware; 9, red, black, and white 
ware; 10, white and green ware. 

Coarse Unpolished Ware 

Although a large collection of coarse ware was made in the excava- 
tions, the forms obtained varied little from those described from 
Awatobi and Sikyatki. Of more than usual interest were specimens 
of eoiled-ware bowls, the interiors, of which were black and glazed. 
These are represented by several specimens from the Chaves pass 

There is but one specimen of rough ware the exterior of which 
was decorated (see figure 17). a The rudeness of the design on this 
object is no doubt in part due to the character of the ware. As we 
go south the number of these specimens of rude coiled ware with 
external decoration increases. They are not found in ruins near the 
inhabited Hopi pueblos, are represented by a few specimens at Homo- 
lobi, increase in number at Chevlon, and are well represented in 
ruins on the northern foothills of the White mountains. 

Undecorated Polished Ware 

While in a few cases polished ware was undecorated, this was excep- 
tional, and only a few specimens were found, which all came from one 
excavation. In certain instances there was evidently formerly an 
ornamentation on some of these which had been obliterated ; on others 
no sign of decoration could be discovered. The polished undecorated 
ware was ordinarily red, but there were likewise specimens of white 
and black undecorated ware. 

Decorated Polished Ware 
red and brown ware 

Red and brown ware is distinctivelv characteristic of the ruins 
found along the Little Colorado, and of those south of this river to 
the border of Mexico. The decorations on this ware (plate xxvi) 
found along the Colorado river are much more complicated than those 
of southern Arizona, where plain red ware is almost universal. In 
both regions the color is no doubt due to the composition of the avail- 
able clay, and to changes in firing. Bricks made from this clay at the 

a A large number of these vessels were found in the more southern ruins excavated in 1897, 
especially that near Snowflake, Arizona, where the largest collections were made that year. 
As this pottery will be discussed at length in the report for that year, it is barely mentioned in 
this section of the memoir. 


present time have much the same texture and color as the ancient 
vessels — probably for the same reason. 


The fine yellow ware which is characteristic of the old ruins near 
the inhabited Hopi pueblos is not found in the Little Colorado ruins. 
There are many pieces (plates xxvn, xxviii) which approach it in 
color, but for the most part they lack that fine gloss which distin- 
guishes ceramic objects of Sikyatki, Old Shuniopovi, and other Hopi 
ruins. While this difference may be in part ascribed to the chemical 
components of the clay, the skill of the potter must also be given due 
credit. While yellow ware was sparingly made in the southern 
pueblos, it reached its highest development in the villages which are 
nearest the modern Hopi. 

In the decline of pottery making the fine old yellow ware has greatly 
deteriorated, and, although clever Hano artists copy it with some 
success, they have never been able to equal the finest specimens which 
the author has dug out of Sikyatki sand hills. 


The Santa Clara pueblo Indians of New Mexico, as is well known, 
make a characteristic black ware. The author has thus far failed to 
find any specimens of this ware in Tusayan, but in the ruins of Homo- 
lobi, Chevlon, and Chaves pass several food basins were found the 
interiors of which were blackened and apparently glazed in the same 
manner as is the Santa Clara pottery. In no instance, however, was 
the external surface thus blackened. Some of these food basins with 
black interior were of coiled ware; others were of smooth ware, but 
all were destitute of other decoration. 


The so-called black and white ware is found almost universally in 
cliff houses throughout the Southwest, and has been thought to be 
characteristic of this kind of dwelling. In his excavations at Sik- 
yatki, however, the author found several pieces, and the same kind 
was also taken from the older quarter of the ruin of Awatobi. Several 
beautiful pieces of black and white ware, with decorations which are 
identical with those of pottery from Colorado cliff houses, were taken 
from the burial places at Old Shumopovi. In the Homolobi and 
Chevlon ruins a number of most interesting bowls, vases, and dip- 
pers of this kind of ware (plate xx) were exhumed, and the same style 
of ware occurred a1 Chaves pass. It appears, therefore, that black 
and white ware is not uncommon in ruins of pueblos in the plains as 
well as in cliff houses, which is but one of manv evidences of tin* 
similarities in culture of the peoples inhabiting these two kinds of 
ancient dwellings. The author was at one time disposed to regard 
these pieces as heirlooms, but the considerable number of specimens 


found would seem to indicate contemporary habitation of the villages 
and cliff houses from which they have been taken. 

No specimen of the black and white ware in the collection is deco- 
rated with designs representing human beings or animals, and even 
pictures of birds, so abundant on other colored ware, are wanting. 
The designs are purely geometrical figures, which are ordinarily 
regarded as the most ancient style of ornamentation. These geomet- 
rical figures, however, are very complicated — as a rule far superior 
to similar decorations on other colored ware. They duplicate for the 
most part the patterns on black and white ware from the cliff dwell- 
ings of southern Colorado and western New Mexico, the headwaters 
of the Salado and Gila rivers. 

Among the specimens of black and white ware there are several 
dippers made of a very fine paste almost as compactly hardened as 
rock. All of these, with one exception, were broken, and the single 
unbroken specimen, one of the most beautiful in the Avhole collection, 
disappeared from the table in the National Museum after it had been 
seen and admired by many visitors. The author mucl^ regrets the 
loss of this beautiful object, especially after it had been brought safely 
to the National Museum. 

While black and white ware is abundant in the cliff houses of the 
San Juan, it is relatively as abundant in the houses of the plains in 
some parts of New Mexico, as may be seen in the great collections 
made in recent times in the Tularosa valley. 


A majority of the ceramic objects from the three ruins investigated 
in 1896 were red with black decorations. This variety was so 
abundant that it may well be styled the characteristic pottery of the 
Little Colorado and its tributaries. Black and red ware is found in 
the ruins near the Zuiii river, an affluent of the Little Colorado, and is 
also found in ruins widely distant from the Colorado, but we are justi- 
fied in regarding this combination of colors as distinctive of the Colo- 
rado drainage area. Some of the best specimens of the glazed ware 
well represented in the collections of 1896 are of these two colors, the 
black designs being almost always glazed. 

The red color is due to the clay, since bricks made at Winslow have 
practically the same color. The many specimens of red and black 
pottery with marginal lines in white on the black form a transition 
from this variety into the next, in which, however, the white is more 


The type of ancient pottery included in the above designation 
(plates xxi, xxn) is, as far as research has thus far gone, peculiar to 
the Little Colorado ruins. No specimen of it has yet been figured 
(1896), and there are no examples of it in the different museums with 
which the author is familiar. 


The three distinctive colors are red, black, and white — the latter 
forming not simply bordering lines in the designs, but being used as 
a slip to cover a considerable surface of the object decorated. While 
specimens of this kind of pottery do not occur in ruins near the 
inhabited Hopi pueblos, it is probable that the modern use of a white 
slip by potters in those villages is a lineal descendant of the ancient 
method of decoration. a 

This colored ware is not found in ruins south of the Mogollon 
mountains, but is confined to the Little Colorado river and its south- 
ern tributaries. 


A limited number of pottery objects of light color, with dark- 
green glazed geometrical figures, were found in the Little Colorado 
ruins (see plate XLiifr). This kind of ware appears to be rare in the 
Hopi country, ancient and modern, but whether it is indigenous or 
intrusive the author has been unable to discover. 


The various forms of pottery are determined largely b3^ the uses for 
which it is intended. They may be classified as follows: 1, food 
bowls; 2, vases; 3, jars; 4, ladles; 5, mugs or dippers; 0, canteens; 
7, cups; 8, animal-shaped vessels; 9, slipper-shaped vessels. 

Food Bowls 

The food basins (plates xxiii-xxix) exceed in number all other 
forms of potter} 7 , and as a rule have the same shapes as those from 
Sikyatki and Awatobi, described in the report on those ruins. 6 The 
basins are ornamented on the interior with symbolic designs, in which 
geometrical figures predominate. 

There is a much larger proportion of designs encircling the exterior 
of the ware in the Little Colorado pottery than in that of Sikyatki, 
and curved lines are also more common. Some of the food bowls 
made of red ware are ver} r large, but from their fragile nature and 
size the majority of these were broken. 

Vases and Jars 

The collection of vases was very large, but the pieces are, as a v 
rule, smaller than those previously described from Sikyatki. Some 
of the forms of these vases may be seen in the accompanying plates 
(xxx-xxxv). The majority are globular, with a slight neck, but 
there are several in which tin' neck is elongated. 

a The use of a white slip is a marked feature of the pottery now manufactured at the East mesa 
of Tusayan. This appears to have been in1 roduced after the fall of Sikyatki, for the fine yellow 
ware of this pueblo shows no white superficial covering. 

b Seventeenth Annual Report of the Bureau of American Ethnology, 1898, pt. :i. 



LETH. ANN. 22 

The designs on vases are usually geometrical; animal and human 
figures are wanting. Some of these vases are very small, having 
evidently been used for pigments or condiments. The lip is some- 
times decorated with parallel marks, but with one exception the 
interior is destitute of ornamentation. A single specimen had an 

Fig. 12. Ladle with divided handle, from Chevlon (number 157051). 

indented or concave base, to secure stability, but in the majority the 
base is rounded. 

The distinction between vases and jars is more or less arbitrary, 
the latter having, as a rule, a smaller orifice. 




Many ladles were found in 1896, the general forms of which were 
in no respect peculiar. The most striking variations are in the form 
of the handles. They are usually decorated on the interiors, and 
generally with geometrical patterns. 

Fig. 13. Ladle with figure on handle, from Chevlon (number 157306). 

Figure 12 shows a ladle witli a double handle — a rare form. Figure 
13 is a ladle with its handle decorated with a human figure. The 
evidences of wear on the edge of this Ladle are pronounced, showing 
that it was probably used in dipping food from bowls or vases. In 
modern times gourd ladles are generally used for drinking purposes. 


Many ladle handles broken from their bowls occur in all the excava- 
tions, and from the appearance of the broken end it is evident that 
the handle was made separate from the bowl and was later joined to it. 
A conical projection from the side of the bowl was inserted into a 
cavity of the handle, which is sometimes hollow throughout, and was 
then luted in place before firing. Several ladle handles were per- 
forated, and in one specimen there were small pebbles which rattled 
when the ladle was shaken. 

The extremity of the ladle handle was at times fashioned into a rude 
image, recalling the clown priests' heads so common on modern clay 
ladles. The ladle with hollow handle opening at one end into the 
bowl, which is so commonly made in modern times by the Pueblo 
potters, a has not yet been found in the ancient habitations of Arizona. 

Fig. 14. Cup with bird designs. 


The canteens by means of which the ancients carried drinking 
water were shaped somewhat like the modern canteen, but were more 
flattened, and generally decorated. While canteens of this shape 
from old Tusayan ruins are known, they have not yet been excavated 
from any of the little Colorado ruins. 

The second kind of canteen, of which several were found, has a very 
different form and probably a different use. While the former was 
generally borne on the back, the second was carried in the hand. It 
has an oval or globular form, with a handle which is hollow, having 
an opening midway in its length. It is possible that this form of 
canteen (plate xxxvi b) was used to carry water for ceremonial pur- 
poses from a spring to the ceremonial chambers, or possibly from 
the houses to the fields. These canteens are of small capacity, and 
are generally ornamented exteriorly with complicated designs. 

"These forms of drinking ladles, made of clay or gourds, are used in the following way: The 
water is dipped up in the bowl and the end of the handle is put in the mouth. A proper slant to 
the bowl allows the liquid to pass through the handle into the mouth. The ancient ladles were 
not used in this way, but were used as dippers are to-day. 

Fig. 15. Dipper from Homolobi (number 




The antiquity of the cup form of household utensil has been ques- 
tioned, but from these excavations there seems no reason to doubt 
that this form was made in prehis- 
toric times. Numerous specimens 
(figures 14-10) were found at 
Homolobi, Chevlon, and Chaves 
pass, and the material of which 
they are made differs in no respect 
from that of other vessels. 

A very fine specimen of coiled 
ware (figure 15) had a handle made 
of two coils of clay artistically 
twisted together. There were a 
few specimens of cups with flat 

bottoms, but the majority were in the form of small vases with 

rounded base. The decora- 
tion of the cups was external; 
the interior was smooth, 
without figures. Geometrical 
figures predominated in or- 
namentation. Several speci- 
mens bore evidences of con- 
siderable use, the rims being 
in some cases much worn. 
One of the best of these cups, 
with handle made of two 
coils, is shown in figure 16. 

The bowl -shaped cup shown 
in figure 17 is made of rough 
coiled ware decorated on the exterior, and has a glossy, black inner 
surface. The form of the handle is exceptional, and is not duplicated 
in the collection. 

Fig. l<i. 3Iug from Homolobi (number 15G891). 

• Pig. 17. Cup, rough ware, decorated, from Chevlon (number 157095). 

A number of vase-form vessels with handles conned the true mugs 
with vases. The general form of these is the same, but the position 

22 eth— 04 5 



[ETH. ANN. 22 

of the handles varies. In figure 18 the handle extends from the lip of 
the vessel to near the equator, while in figure 19 it is smaller, and 
placed just below the neck of the vase. In figure 17 the handle is 
confined to the equatorial region. 

— i 


Fig. 18. Mug. 

There were several specimens of a mug form in which the body is 
trifid. These forms were probably used for pigments or condiments, 
and were of rough ware, or were polished and sparingly decorated. 
The external decoration of these cups varies in character as widely 

as does their form ; but in 
plate xxxi v the reader will 
find some of the designs, 
which are practically the 
same as those on the in- 
side of certain food bowls 
from the same ruins. 

Animal Forms 

Pottery objects in the 
forms of birds, though 
common among the ce- 
ramic productions of cer- 
tain modern pueblos, are 
rarely found in ancient 
ruins. The excavations 
made in 1896 brought several specimens of these to light, one of the 
best of which was from Chevlon (figure 20). In this specimen we 
have a well-made head recalling that of a duck, and three knobs 
representing the tail and wings. Interesting in connection with this 
specimen is the presence of triangular designs with terraced figures 
painted on the sides. 

It may be supposed that the vase (figure 21), with four knobs 

Fig. 19. Mug from Chevlon (number 1572U4). 




arranged at equal intervals about the equator, is a highly conven- 
tionalized bird vase in which the head, wings, and tail are represented 
by knobs or rounded elevations. Figure 20 has on one of the knobs a 
head with goggle eyes and teeth, recalling a bird figure which is not 

Fig. 20. Duck-shaped vessel from Chevlon (number 157018). 

an anomalous form in southwestern pictography. In view of the 
identification of the terraced figures with wings, the decoration on 
the equator of this bird-effigy vase are highly suggestive. 

A second bird-form vase is shown in figure 22, in which the vessel 


Ik;. 21. Jar with four knobs, from Homolobi (number 156364). 

is elongated and has a round head projecting at one end, with a beak 
like that of a parrot . The eyes project from the head, and there is an 
opening for a mouth. On the side of the body there; are parallel lines 
representing feathers painted in red and black. The tail is broken. 



[ETH. ANN. 22 

A third kind of bird-form vase is more globular in shape, with head 
appended to the rim. The wings and tail are represented by figures 
drawn on each side of this vase, and eyes are painted on the sides of 
the head. The strange raised S-shaped bodies on the sides of this 
vase are of unknown meaning, but they suggest legs. A view of 

Fig. 22. Bird-shaped vessel from Chevlon (number 157909). 

this vase from one side is shown in figure 23. The length of the 
appendages represented in relief on this vessel suggests some long- 
legged wading bird, possibly a crane. 

Plate xxxvi a shows a conventional effigy bowl in which the rim is 
modified to represent the head, tail, and wings of a bird. 

Fig. 23. Bird-snake vase from Chevlon (number 157311 ). 

The preceding forms, representing all effigy objects which were 
found in the excavations, naturally lead to a consideration of the great 
difference in the ceramic technic of northern and southern Arizona 
and New Mexico. As we go south there is an ever-increasing tend- 
ency to combine relief with pictorial decoration; effigy vases, includ- 
ing those in the form of men and animals, increase in number, until 


in Mexico relief decoration becomes the essential feature. In its 
earliest development the head is raised in relief, while arms and legs 
are indicated by ridges, as in a figure from the Nantacks, considered 
later in this article. In the Gila valley, clay reliefs of the human 
figure on a jar or vase Avere attempted — a thing unknown in ancient 

Slipper-shaped Vessels 

Several rough-ware jars in the form of slippers were found in the 
ruins excavated in 1896. From the fact that many of these were 
blackened with soot, it is conjectured that they were formerly 
used for cooking vessels, and it is probable that they were made in 
that peculiar form in order that they might be used like Dutch ovens 
and coals of fire might more readily be heaped over them. Many of 
these slipper- shaped jars had one or more handles placed on the necks 
or prolongations of the rims. 

These jars were always made of rougn ware, and were never 
painted, as is the case with similarly formed vessels from the Gila 
river and its northern tributaries. They vary in length from a few 
inches to a foot or a foot and a half. 

General Character 

The great value of collections of pottery from the Southwest, espe- 
ciallv from the ruins in Arizona, is to be found in the symbolic deco- 
ration and its interpretation. The collections in 1896 were especially 
instinctive on account of the new localities from which they were 
made and of the new sjnnbols depicted. As is universally the case, 
avian figures are the most common and the most elaborately con- 
ventionalized. There are one or two instructive reptilian designs. 

A study of the decoration on the pottery of the Homolobi, Chevlon, 
and ('haves pass ruins shows that the proportion of geometrical to 
animal designs is much larger than at the Sikyatki or Shumopovi. 
In the few instances where animals and human beings are depicted 
the execution of the designs is ruder. This preponderance of geomet- 
rical over animal figures recalls conditions characteristic of white and 
black ware ornamentation. The predominance of animal pictographs 
on pueblo pottery in ancient times appears especially characteristic 
of Tusayan. 

The most novel results obtained from a study of the collections of 
poll cry wen 1 contributions to a knowledge of ancient pictography. a 

Even a superficial comparison of the pictography of the Little Colo- 
rado pottery with that of tin? Sikyatki ware shows how inferior the 

"The majority of forms of ancient Tusayan ware are well known to archeologists through 
the Keam collection, some of the more striking specimens of which have been figured by Mr 
Holmes in previous reports of the Bureau of Ethnology. 



[ETH. ANN. 22 




Fio. 24. Footprints on inside of a vase 
from Homolobi (number 156690). 

former is to the latter. A deeper 
study confirms this conclusion. No- 
where in the pueblo region has the 
-« ceramic art reached the high position 
it attained in Tusayan in prehistoric 
times. This development may be due 
in part to the character of the material 
used, but it is mainly owing to the 
artistic instinct of the ancient Hopis. 
In both the character of the paleog- 
raphy and the texture the ware of the 
Little Colorado ruins is more like Zuni 
than Hopi work. The ancient pueblos 
on the Zuni river, a tributary of the 
Little Colorado, closely resemble those 
about Winslow and at the mouth of 
Chevlon fork, but their pottery is as 
a rule inferior. 

Human Figures 

Pictures of human beings were very 
rarely found in the excavations at 
the Little Colorado ruins. This rarity 
conforms with results from other ruins, 
described by other archeologists, so 
that the author suspects that delinea- 
tions of the human figure, of which 
several were found at Sikyatki, indi- 
cate a late stage in the evolution of 
pottery decoration in ancient pueblos. 
The drawings of human beings which 
have been found are for the most part 
of the rudest possible character, show- 
ing no elaboration such as would be 
expected if they had been used many 
generations for decoration. 

But a single complete figure of a 
human being on pottery was ex- 
humed in 1896, and that was on the 
handle of a ladle from Chevlon. The 
specimen (figure 13) represents a 
woman with left arm lifted high above 
the head. It has been identified as 
the figure of a woman from the pres- 
ence of the characteristic coiffure of 



maidens, to which the author has called attention in his account of 
designs from Sikyatki. The end of the handle of this ladle turns at 
right angles, and suggests an explanation for numerous clay objects 
of like shape which have been found elsewhere. 

Among human figures, however, may be mentioned the unique 
ornamentation on the inside of the vase from Homolobi (see figure 24), 
where we have on one edge the representation of the head, neck, and 
extended arms, one of which carries a rattle, the other a spear. A 
line of footprints extends across the inner surface of this vase, and 
the body and legs are represented on the opposite side. It will be 
noticed that the portions of the human figure represented at the two 
ends of the line of footprints are complemental ; the head and arms 
appear at the bottom, the body and legs at the top. It would seem that 

Ftg. 25. Quadruped figure on food bowl from Chaves pass (number 157570). 

the artist intended to represent the tracks of a seated figure at the 
bottom of tin* line of footprints, the marks being paired at that point. 
The inside of the food bowl shown in plate xxvii b is decorated 
with a human face, in which eyes and mouth are represented. Above 
the head is a crescentic figure in white, resembling the moon, into the 
concave side of which project four pairs of tubercles from the top of 
the head. This is one of .the few specimens from Homolobi in which 
a human face is depicted. 

Quadruped Figures 

There were a few pictographs of four-legged animals, two of which 
are identified as mammalian forms. Mythical lizards and batrachians 
are represented, but no complete picture of any reptile was found 
which could be identified. 



[ETH. ANN. 22 

One of the most interesting pictures of mammalian animals occurs 
on a broken food basin from Chaves pass. This specimen (figure 
25) represents an animal with long claws, a tail reaching above the 
body to the head, two triangular ears, and an arrow-shaped tongue. 
It is one of the few figures in which the intestinal tract is represented, 
and it has two eyes on one side of the head. 

The design on the food bowl shown in figure 26 represents a four- 
footed animal which was identified by one of the Hopis as a bison, 
and the hump on the back certainly suggests this animal. This 
figure, like the preceding, has two eyes on one side of the head, but, 
unlike it, has the four legs all depicted in the same plane. The 

Ftg. 26. Quadruped figure on food bowl from Chevlon (number 157102). 

geometrical figures below this quadruped are of unknown meaning. 
The bowl is of red ware, with black and white decoration, and is one 
of the finest of this kind from the Chevlon ruin. 

As a rule, vases are ornamented on the equator, and it rarely 
happens that any design is found on the bottom. The specimen 
shown in figure 27, however, has a design in that region resembling 
a paw of some animal, possibly a bear or badger. The form and char- 
acter of ware which distinguish this specimen are likewise highly 


Bird Figures 

The majority of the animal figures on specimens from the three 
southern ruins represented birds, many of which were highly con- 
ventionalized. While there were many objects of pottery adorned 
with feathers, this stjde of decoration was not as common or as 
varied as at Shumopovi, Sikyatki, Awatobi, or other ruins on the 
Hopi reservation. The conventional forms of feathers so common 
on the decorated pottery of Sikyatki are not found in the designs 
ornamenting the pottery of the Little Colorado ruins, but seem to be 
confined to the pueblos in the present Hopi reservation. Thus, not 
a single specimen of the conventional feather figured on the "butter- 
fly vase" shown in plate cxxv of the Seventeenth Annual Report of 
the Bureau of American Ethnology, part 2 (and also plate XL, Smith- 
sonian Report, 1895), was found on any vessel from Homolobi, Chevlon, 
or Chaves pass. a 

Fig. 27. Vase with bear's paw design (number 157187). 

The peculiar symbol of the breath feather (Seventeenth Annual 
Report of the Bureau of American Ethnology, plates cxxxviii b and 
cxli c, (1) also appears to be limited to objects from ruins near the 
inhabited Hopi pueblos. On none of the many figures of birds shown 
in the Little Colorado pottery have we any such complicated symbols 
appended to wing or tail. The figures of birds from Shumopovi 
resemble those from Sikyatki, but no pottery from a Little Colorado 
ruin is found decorated with the conventional figure of the feather so 
constant in the 1 ancient rains above mentioned. 

It will be noticed in t he figures of birds from Homolobi and Chev- 
lon that the posterior end of the body has a triangular form which 
apparently represents the tail. At one side of this triangular figure 
are many short parallel lines, evidently intended to represent the 
tips of the tail feathers, well brought out in the bird figures. 

The design shown in figure 28 represents two birds, above which 
are emblematic rain-cloud symbols with parallel lines representing 

"Consult The Feather Symbol in Ancient Hopi Designs, American Anthropologist, v. 11, n. 1, 
January, 1898. 



[ETH. ANN. 22 

falling rain. The figures of these two birds are decidedly Egyptian 
in form. Their beaks are turned in the same direction, and both 
have two eyes on one side of the head. The wings are of special 
interest to students of Hopf symbolism, for they are represented by 
triangles — which is often the case in ancient Tusayan pictures 
of mythic birds. This bowl, found at Chevlon, is of the red 
ware characteristic of the Little Colorado ruins. The triangles at 
the lower ends of the bird figures are tails, and the short parallel 
knobs represent the tips of the tail feathers. This is important to 

Fig. 28. Mythic bird figures and rain-cloud symbols on food bowl frorn Chevlon 

(number 157221). 

remember in the study of symbolism, for we sometimes find the same 
symbols depicted alone on a vase (see figure 36). 

No specimen from the Little Colorado ruins has a diametrical line 
representing a ''sky band" to which hangs the conventional figure of 
a bird — a design so common in the best Sikyatki ware. A good exam- 
ple of this ornamentation is shown in a food bowl from Shumopovi 
(see figure 73, page 117). We miss also the star design and the 
trifid cross so commonly associated in Tusayan ware with the bird 




; "-T-v. 



Attention is called to the form of the tail of the two birds in figure 
28 and to the triangular de- m 

signs called feathers seen 
in the same figure. It 
seems not improbable that 
in the conventionalization 
of bird figures the design 
representing a bird may 
be reduced to two trian- 
gles, making an hourglass- 
shaped figure. Suppose, for 
instance, wings and head 
be omitted in figure 28, the 
tail and body would then be 
two triangles joined at the 

The design on a vase from 
Homolobi shown in figure 29 represents four birds, each one of which 


Fig. 29. Vase with four bird figures, from Homo- 
lobi (number 156676). 

I'n. 30. Mythic bird figure on food bowl from Chaves pass (number L57563). 

has a crested head of feathers and widespread wings of triangular 

form. The body is continued into two triangular extensions, as is 



[ETH. ANN. 22 

the case in so many bird figures, and the tail feathers are indicated 
b} T short, stumpy, parallel lines attached to one side of a triangle. 
The middle of the body is represented by a lozenge-shaped figure, in 
the center of which is a dot. Trifid triangular designs alternate 
with the bird figures, and the bird figures are arranged as though 
moving in a sinistral circuit. 

The figure of a bird on a food bowl from Chaves pass (figure 30) is 
characteristic. This represents a toothed bird, a conception often 
repeated in the ancient pueblo pictography. Both eyes are on one side 
of the head, which is rounded posteriority and prolonged into an upper 
and lower dentate beak. The triangular wings are terraced or 

Fig. 31. Bird design on food bowl from Homolobi (number 156603). 

notched on one edge, and the tail is triangular, with short, white 
appendages representing feathers. Although a simple figure, this is 
one of the most instructive bird designs in the collection. The con- 
ception of a toothed bird is certainly remarkable, but we find it still 
current in the AValpi ritual, where it is personated, as in the so-called 
Natackas which appear in the Powamu, or Bean-planting, a ceremony 
when the fields are prepared for planting. 

The figure of a bird represented on a food bowl from Homolobi 
shown in figure 31 is different from any elsewhere collected. Par- 




tieularly interesting is the drawing of the wing and the shape of the 
body, which is bordered by small triangles. Both eyes are repre- 
sented on one side of the head, and the tail feathers, four in number, 
are represented in a vertical plane. 

The food bowl shown in figure 32, from Homolobi, is decorated on 
the interior with a design representing the head, neck, and legs of a 
mythic bird. There are two eves on one side of the- head, and the 
tongue has a tip like an arrowhead. The wings bear triangular 
appendages representing feathers. The talons recall those of the 

Fi<;. 32. Bird figure on food bowl from Homolobi (number 156870). 

"unknown reptile' from Sikyatki, figured in plate LXI] of a pre- 
liminary report on 1 hat ruin/' We have in this figure a representation 
of both wings in the same plane, a constant feature in Pueblo draw- 
ing. There is also a view of a body cavity, which is not rare in 
modern Pneblo figures of animals. 

One of the most striking pictures of birds is that depicted on the 
interior of the food bowl from Chevlon, shown in figure 33. The 

"Annual Report of th<* Smithsonian Institution for 1 s«.»r>. 



[ETH. ANN. 22 

most remarkable of all the appendages are those on the tail, the mean- 
ing of which the author can not interpret. It was sometimes cus- 
tomary to equip a bird figure with a long snout in which were teeth, 
and this conception persists among the Hopis, as has been noted above. 
It is interesting to note that in this figure, as in the majority of bird 
figures from the Little Colorado ruins, the tail is represented by a 
triangle, and the tail feathers or their tips by three parallel lines. 

The interior of the food bowl shown in figure 34 is decorated with 
a bird design which exhibits some of the notable violations of per- 
spective common in ancient Tusayan art. We here find wings, legs, 

Fig. 33. Mythic bird figure on food bowl from Chevlon (number 157264). 

and tail feathers shown on the same plane, notwithstanding that a 
side view was intended. 

The indication of the claws by crescents in this figure is interesting. 
The same method is adopted in another bird figure, in which there 
are in each foot two short parallel lines. This method is likewise 
used in one of the designs from Sikyatki which was identified as rep- 
resenting an unknown reptile. a There is some doubt whether this 
figure represents a lizard or a bird, for a considerable part of the 
body is posterior to the appendages. If we consider the posterior 

« Seventeenth Annual Report Bureau of American Ethnology, pt. 2, 1898, figure 269. 




appendages as a pair of legs, they bend the wrong way, unless the 
whole portion from the angle to the claws is regarded as foot. This is 
not an avian feature, but the presence of semicircles and triangles on 
the body is characteristic of bird symbolism. 

In studying the different figures of reptiles from ancient pottery the 
author finds no other in which the feet have this form — which occurs 
in undoubted bird figures from Homolobi and Chevlon. It might 
therefore be concluded that the Sikyatki figure was wrongly identi- 
fied and should be called a bird. There are, however, almost fatal 
objections to this identification. The most striking of these is the 

Pia. -tt. Bird figure on food howl from Chevlon (number 157084 >. 

elongated form of the body. The anterior appendage, which is 
identified as a leg, can hardly be homologized with a wing, although 
it must be confessed that the parallel lines may be feathers. 

The reexamination of the figure in the light shed on the subject 
by the bird figures from Homolobi reveals that it has both bird and 
reptilian features, and that the former predominate. 

Two raptorial birds are painted on the food bowl shown in figure 35, 
on opposite sides of terraced figures which recall cloud symbols. The 
birds are shown in profile, with both eyes on one side of the head 



[ETH. ANN. 22 

and tail feathers thrown out of perspective. Wings are not repre- 
sented, and the body is covered with cross-hatched lines. In a bowl 
from Shumopoviwe likewise, find two birds represented from a differ- 
ent point of view, and also terraced figures which have been inter- 
preted as rain-cloud symbols. 

The triangular designs on the vase shown in figure 36 are inter- 
preted as feathers, or rather as the tails of birds with appended 
feathers. The reason for the interpretation is to be found in the 
study of the bodies of birds as represented in the Little Colorado pot- 
tery. The author has elsewhere shown seA^eral instances in Sikyatki 

Fig. 35. Food bowl with bird designs. 

pottery where symbolic feathers are represented as tied about the 
neck of small vases, and even at the' present day certain gourds in 
which sacred water is brought from springs in kiva ceremonials have 
feathers tied in this position. The triangle, as a feather symbol, is 
still found in certain altar pictures — as the snake-lightning designs in 
sand of the Antelope altars. It is interesting to notice that these 
feather symbols have parallel white lines on one side. 

The only other symbol with which these triangular figures on the 
Homolobi vase could be identified are triangular rain clouds, the 
short parallel lines representing falling rain. These symbols would 




also be appropriate on this small vase, but there is more likelihood 
that the triangles in this instance are feather symbols. 

Insect Figures 

In an account of the insects used in pottery decoration at Sikyatki 
attention was called to the use of the moth or butterfly and the 
dragon fly. Both of these forms occur on pottery from the ruins 
along the Little Colorado, and their symbolism appears to be the 
same in all the ruins in Arizona thus far studied. The butterfly is 
commonly indicated by a triangular figure, which often becomes 
highly conventionalized, as in plate xxv b. 

Pig. 36. Vase with bird symbols, from Homolobi (number L56880). 

Arachnid Figures 

The spider plays an important pari in Pueblo mythology, and the 
so-called Spider woman is often mentioned in connection with the Sun 
and the war god. 

The design on the food bowl from Homolobi shown in figure 37 
represents a spider, and a figure of the sun on the outside of this 
bowl recalls the legend of the Spider woman who married the Sun. 
The association of these two symbols on an ancient vessel shows 
the antiquity of this well-known legend. 

22 eth— 04 6 



[ETH. ANN. 22 

There are appended to the cephalotliorax of this animal four pairs of 
legs, which number distinguishes arachnids from insects; there are 
two jaws at the anterior and a feather at the posterior extremity. 

The figure of the sun on the exterior of this food bowl is a simple 
ring surrounding a white zone, in the interior of which is a black 
spot. The four peripherally placed sets of three parallel lines are 
supposed to represent eagle feathers, a constant feature in sun 
emblems, or red horsehair, symbolizing the rays of the sun. The 
sand picture of the sun in the Powalawu, a ceremony preceding the 

Fig. 37. Spider and sun emblem on food bowl from Homolobi (number 156888). 

Powamu in February, as made in Oraibi, is identical with the design 
on this food basin. 

Kokyan wiiqti, or as she is generally called, Kokyan mana, the 
Spider maid, was the parthenogenetic mother or grandmother of the 
twin Avar gods, Puiikon hoya and Paliina ho3^a. As she was supposed 
to have been impregnated by a sun's ray or a drop of water falling 
upon her, the sun is therefore called the father of the twins. 

In various eurrent legends the Spider woman is represented as one 
who can change her form at will, acting as mentor to the hero Puiikon, 



generally perching on his ear, and whispering her promptings from 
that position. She assumes several roles and is designated by many 
attributal names. She is sometimes called "creator," but is an 
earth goddess or mother rather than an artificer of nature. 

Geometrical Designs 

The wealth of geometrical ornamentation (plates xxvm, xxix) in 
Pueblo pottery decoration has attracted the attention of many stu- 

FlG. W. Three lines of life. Design on food bowl from Chevlon (mimber 156138). 

dents, and a large literature lias accumulated on that subject. This 
form of decoration is the simplest in motive, the most widely spread over 
The pueblo area, and also probably the most ancient. Its relations to 
decoration of textile art products arc pointed out by several writers, 
and there arc many evidences of the evolution of complicated geo- 
metrical figures from simple forms. There are also evidences of their 
origin by conventionalization of more elaborate patterns through 

The design on the food basin shown in figure 08 is unique and highly 
instructive in one important feature. Encircling hands or lines on 
specimens of ancient pottery are ordinarily broken at one point, as 



[ETH. ANN. 22 

can readily be seen by an examination of figures in the report on 
Sikyatki pottery, as well as in the present memoir. The design before 
us has three breaks in these encircling bands. The break in the 
interior band is complicated by the addition of well-known terraced 
figures. By modification in form and by the interlocking of these 
appendages we pass easily to some of the most complicated geomet- 
rical patterns of Pueblo pottery. 

A modification of the broken line about a vase appears in the 
specimen shown in plate xxxvic. In this instance we find the band 
continued into two narrow extensions, which interlock but do not 

Fig. 39. Geometrical designs on a food bowl from Chaves pass (number 157539). 

join. The figure which is thus formed is a common one in geomet- 
rical ornamentation, as may be seen by an examination of the many 
beautiful pieces of pottery obtained from the ruins in all parts of 

Not less instructive than the last-mentioned in a study of geomet- 
rical ornamentation is the design in figure 30. The spiral figures on 
two of the bands show a modification of the broken lines which are 
characteristic, and the S-shaped ornaments on one of the other bands 
are common on ancient pueblo ware. 




A combination of oblique lines and fret, shown in figure 40, is the 
most unusual design in the collections which it luis been the author's 
fortune thus far to examine. The same design is worked in a basket 
from the same ruin (figure 63). 

The design figured on the interior of the food bowl shown in fig- 
ure41 is unique, 
reminding one 
of a swastika 
with arms split 
into two parts. 
We recognize in 
it the familiar 

Fig. 40. Broken 
fret on food bowl 
from Chevlon 
(number 157896 I. 

Ftg. 41. Food bowl with geometrical patterns (number 156427). 

triangle and crook, but the long scythe-like projec- 
tions on the periphery of the design are rarety found 
in ancient pottery decoration. 

The decoration of another food bowl from Ilomo- 
lobi is likewise unique, but it is formed of la miliar 
decorative designs arranged in an irregular manner. 



The ancient pueblo peoples of Arizona were adepts 
in making mosaic, some examples of which rival in 
excellence the work of a similar kind in old Mexico. The author's 
explorations in L896 revealed different kinds of this craft, several of 
which are unique. The southernmost ruin, that in Chaves pass, 
yielded the most beautiful specimen, but those found at Chevlon 
were almost as fine, and the variety was greater. 



[ETH. ANN. 22 

Fig. 42. Mosaic gorget from 
Chaves pass (number 157850). 
Length about 2 inches. 

In the course of excavation there were found in the soil, near a 
buried body, a large number of accurately squared turquoises which 
were so small that they could hardly have served for ornament alone, 
and with them were found worked shells covered with a tenacious 
pitch. In one of these collections there are several larger fragments, 
evidently turquoise ear pendants. Later excavations explained the 

character of these turquoises, for they were 
found to be duplicated in specimens of fine 
mosaic ware. 

It is well known that Hopi women at the 
present day wear ear pendants made of 
square wooden plates, upon which are ce- 
mented rude mosaics of turquoise. The modern work of this kind is 
comparatively coarse, and evidently is made of old turquoises, some 
of which are perforated and were formerly used as beads. The tur- 
quoise stones employed are not accurately fitted, and the black gum 
in which they are embedded shows between the stones. The ancient 
work (plate xliv) is much finer and more beautiful than the modern. 
Specimen 159850 is a turquoise mosaic set on wood, but it was so 
broken that it was impossible to tell what its form was. 

The specimen shown in figure 42 is an elongated gorget of wood 
with shell and turquoise incrustation; it was 
found at the Chevlon ruin, and is one of the 
most beautiful of prehistoric mosaics. Figure 
43 shows a pear-shaped pendant made of bone 
with turquoise mosaic on one side and incised 
chevrons on the opposite. It is perforated at 
the blunter end as though for suspension. 
The specimen was taken from the Chevlon 
ruin, and is unique. The collection contains 
also a number of fragments of turquoise and 
other stone mosaics, and of catlinite and tur- 
quoise embedded in gum on wood. These were 
from the Chevlon ruin. Many other square 
turquoises, evidently formerly parts of a mosaic, were collected at the 
same ruin. 

The specimen shown in figure 43 was found near the mastoid process 
of the skull of a woman. It was evidently an ear pendant, one 
attached to the ear by a string. Two other specimens of bone incrusted 
with stone mosaic were found at Chevlon. 

One of the most beautiful examples of mosaic was a worked shell of 
Pectunculus giganteus covered with turquoise stones embedded in 
pitch. It was found on the sternum of a skeleton from Chaves 
pass, and was buried several feet below the surface of the ground. 
Plate xliv shows two views of this unique and precious specimen, 
one from above and one from the umbo of the shell. In the former 

Pig. 43. 




Bone ear pendants 

Chevlon (number 

Length about 2 




a median rectangular fragment of red jasper is shown, and in the 
latter appears the hole by which the ornament was formerly sus- 
pended. The latter likewise shows legs, suggesting a frog, turtle, 
or toad. The arrangement of the lines of turquoise on the opposite 
rim, obscurely seen in the upper view, has been regarded as represent- 
ing the hind legs of the same animal. The technic of this mosaic 
recalls work of the same general character on dirk handles and masks 
from Mexico. 

Several additional specimens of similar stone mosaic on shells were 
found, but these were much broken and impossible of restoration. 
The mosaic frog was broken when found, but the anterior end was 
entire and still clinging to the shell when dug from the ground. 

A summary of the specimens of mosaic collected is given with their 
catalogue numbers in the following list: 

Number Article 

(1740) Squared fragments of turquoise formerly incrusted on shell 

157849 Square lignite pendant for ear, with turquoise in each corner 

and middle 

159850 Pendant of wood incrusted with stones 

157852 Pyriform ear pendant of bone, with turquoise incrustation 

158068 Fragments of a mosaic of turquoise and other stones 

157848 Fragments of stone mosaic on wood 

157932 Fragments of stone mosaic on wood 

157851 Mosaic frog. toad, or turtle 


One of the common uses of lignite was that of ornament. The 
specimen represented in figure 44 is a square 
ear pendant, taken from near the mastoid 
process of a human skull in the Chevlon 
cemetery. It is made of lignite, nicety pol- 
ished and accurately squared. In each cor- 
ner and in 1 he middle on one face irregular 
turquoises are sot in depressions, while on 
the obverse, near the middle, there is an 
cy<'l<'t in the substance of the pendant. 
The hut ton-shaped gorget shown in figure 
45, unfortunately broken, was flat on one 
side, and convex on the opposite, which was 
smoothly polished. The stria' across the 
flat side suggest the polishing stone, and 
the perforations point to suspension by a 
string or thong. This is one of the best formed lignite buttons in the 
collection, but ruder forms have been taken from other ruins. 

FiG. 44. Lignite ear pendant 
from Chevlon (number L57849 > . 
Natural size. 



[KTH. ANN. 22 

Unworked fragments of lignite are rare, and the material appears 
to have been brought* to Chevlon from some distance, although it is 
common in the rocks near the modern Hopi villages. 

Fig. 45. Lignite gorget. Slightly reduced. 


The collections made in the summer of 1896 were particularly rich 
in ornaments made from marine shells. The largest number of these 
were found in the ruins at Chevlon and Chaves pass, although a con- 
siderable number of specimens were collected from the ruins of the 
Homolobi group. 

The shells used in making these ornaments belong to the Pacific 
coast fauna, and no doubt came through barter to the people 
who once inhabited the towns of the Little Colorado, for it is well 
known that there was a considerable trade in early times in these 
shells, and long trips were taken by the Pueblo Indians for trade 

The intercourse of northern and southern peoples of Arizona 
through trading expeditions continued to quite recent times, but 
judging from the number of specimens which were found in 
the ruins it must have been considerably greater in prehistoric 
times than it is at present. In fact, much of the decline in this 
traffic is probably to be traced to the modification of the southern 
Arizonian aborigines and the introduction of new ornaments by the 

One of the most highly prized of these marine shells was a species 




46. Incised armlet from. Chevlon (number 
157843). Diameter about 3^ inches. 

of Pectunculus," which was worked in many shapes, or preserved in 

practically the same form as when taken from the sea. A number 
of these were very fresh looking; others were more or less decayed. 

Some of the most characteristic specimens are shells witli round 
holes in the middle. It would appear that these are ornaments; and 
as one of these objects was found near the wrist bone of a man it was 
judged to be a wristlet. It is suggestive that these perforated shells 
were generally found in pairs, as though belonging to some dual 

Five specimens of perforated Pectunculus shells were found at 
Chaves pass (one of them a 
fragment smeared with gum), 
and the same number at Ilomo- 

Pectunculus shells were 
worked into armlets, or large 
rings, 6 by cutting out the whole 
middle of the shell, leaving the 
rim, which was ground to a 
smooth surface. The umbo 
was carved into a shell-shaped 

elevation, and the surface was often incised with characteristic designs. 
The following specimens were found: 

Number Locality and description 

157824 Chaves pass: 3 specimens 

157659 Chaves pass; 2 specimens 

157704 Chaves x^ass: found on a humerus 

157205 Chevlon: with inserted turquoise 

157843 Chevlon: with incised design 

156760 Homolobi: 2 specimens 

157002 Homolobi: fragments 

There are also several fragments from an unknown locality. 

As a rale, the surface of these armlets is smooth and without 
ornament, but one specimen (figure 40) was beautifully decorated with 
a characteristic incised fret covering the whole outer surface. The 
design consisted of a series of lines interlocking at extremities, though 
not joining. This figure is one of the simple forms of a characteristic 
decorative motive widespread over the whole pueblo area. In its 
simplest expression it appears as two crescents turned in opposite 
directions, with the two horns adjacent. It is thus painted on lli< i 
breasts of certain katcinas (personations of supernatural beings), and 

"The following specimens were taken from Homolobi, Chevlon, and Chaves pass: Number of 
specimens. 114: inCTOSted with stone mosaic, 1; incrosted with pitch. 1: armlet, incised. ]; arm- 
let, inlaid with turquoise, 1: armlets, not ornamented, 10 (many additional fragments); wristlets, 
44 (many additional fragments); finger rings, '■#) (many additional fragments); fragments 
incrustedon wood. 2: carved in imitation of frog, 1: shells not worked,:}; shells with medial 
perforation. 20. 

& Popularly but erroneously called "earrings.' 1 



[eth. a nx. 22 

on shields, or is cut in pictographs. But it is in decoration of pottery 
that this simple form reaches its highest modification and complica- 
tion, and it is remarkable how many complex figures can be reduced 
to this simple type. The horns of the two crescents may elongate and 
develop into square frets or spiral extensions, and these in turn may 
be continued into triangular appendages with dentate or serrate mar- 
gins. They may become terraced figures, their edges so closely approxi- 
mated as to be separated by zigzag intervals, which in all cases are 
but the space left by the break. With all these modifications, no 
matter how complicated, the motive can be reduced to the two horns 
of adjacent crescents opening toward each other, but not joining. 
The break is comparable to that in encircling bands drawn on pueblo 
pottery, called the broken lines or "lines of life.' 1 Consider such a 
line about a vase, bowl, or jar to be broken at several points, or, as 
the author found in several instances, to have three breaks, and the 
ends so extended as to overlap the intervals either above or below; 
modify the extremities thus extended into terraced figures, spirals, or 

frets, and we have some of the 
developments of this most char- 
acteristic of all motives in the 
geometrical designs of decorated 
Pueblo pottery. This broken 
line, with its modifications, is 
used almost universally as a 
decorative motive by Pueblo 
potters, ancient and modern, 
whether living in pueblos, cliffs, 
or caves. The design on the 
armlet shown in the figure is a modification of the same motive. 

A single specimen of armlet, shown in figure 47, has a turquoise set 
in the outer surface near the edge. The stone was probably fastened 
there with pitch, the armlet being the only specimen of shell inlaid 
with turquoise in the collection. 

The wristlets were made of the same genus as were the armlets, 
but from smaller specimens. A number of these ornaments were 
found in some instances encircling the radius and ulna. The major- 
ity were from Chaves pass. Twenty-nine complete specimens were 
secured here, together with many fragments, and one specimen was 
found at Homolobi. 

Bracelets made of this shell are smaller, slighter, as a rule less care- 
fully worked, but more abundant than the armlets. The majority 
are perforated at the umbo, but the valves are so ground down that 
there remains no space for the heart-shaped elevation; indeed, the 
thickness of the shell would not admit of it. Like the armlets, they 
are sometimes found free in a grave, as though cast there as a votive 
offering, but there can be no doubt that they were bracelets, for in 

Fig. 47. Armlet with inlaid turquoise, from 
Chevlon (number 157295). Diameter about 3^ 




more than one instance the author has taken them from the bones of 
the arm. Similar specimens often have been described as ear pend- 
ants, and they may in some instances have served for this purpose, 
but all of these objects found in the Little Colorado river graves were 

A number of finger rings made from small specimens of Pectun- 
culus were found at Chaves pass. 

Two fragments of Pectunculus were fashioned in the shape of gor- 
gets. One of these, from Chevlon, was perforated with two holes; 
the other Avas crescentic, with a single perforation. The latter was 
found in a grave at Chevlon. 

Unworked specimens of Pectunculus were not numerous, but three 
were found at the Chaves pass ruins, and one at Chevlon. In this 
connection may likewise be mentioned a clay 
imitation of a Pectunculus shell from Kisakobi, 
the site of old Walpi. 

Two specimens from Chaves pass were 
smeared with a tenacious pitch and probably 
f ormerly incrusted with turquoise, squared frag- 
ments of which Avere found near by. 

Three species of Conus (C. fergusoni, C. 
princeps, and C. regularis) were found in pre- 
historic graves. These were favorite shells for 
the manufacture of rattles, and they are still 
used for that purpose by the Hopis. The spire 
was ground away on a plane at right angles to 
the lip, making a conical object perforated at 
the apex. The larger specimens (see figure 48) 
were probably tied to a short crook, and were used as rattles with 
which to beat time to the sacred songs. Smaller specimens, found in 
great numbers on some of the skeletons, served as tinklers, and were 
apparently tied to garments of the deceased in much the same fashion 
that the tin cones are appended to the kilts of Snake priests in the 
Snake < lance. 

The number of specimens of Conus, especially of Conus princeps, 
was large, and the majority were found at the ruin in Chaves pass. 
One of the best preserved of the specimens is shown in figure 48. 
This was probably a part of an ancient rattle, and the cut shows the 
hole which formerly served as the place of suspension. Several of 
the sum Her specimens were found near the pelvis of a skeleton, as 
though they were formerly tied to kilts, as is suggested above. 

A single specimen of the Turritella shell was found at the Chevlon 
ruin, but with the exception of a perforation near the lip it was not 
worked. At the present day Turritella is so highly esteemed that 
specimens of it are attached by a string to several of the Upon is, or 
palladia, of religious societies. 

Fig. 48. Shell used for 
rattle, from Chevlon 
(number 157847). 



[ETH. ANN. 22 

Fig. 49. Shell ornament from 
Homolobi (number 156391). 

Haliotis shells were prized by the ancient Hopi pueblos, and are 
still highly regarded and used for decoration among the modern 
Tusayan Indians. Three specimens were found at Chaves pass. 
These were entire, though very much eroded, when they were 

dug out of the earth. They were the larg- 
est and most beautiful specimens of Hali- 
otis which the author has seen from ancient 
Arizona ruins. Several fragments of this 
shell were . found, all apparently worked, 
two being perforated for suspension. 

A Strombus shell from the Chevlon ruin 
has a ring of pigment about the umbo, but 
one from Chaves pass is undecorated. 

One of the most highly prized for orna- 
mental purposes of all Pacific coast shells 
was Cardium, which made its way by bar- 
ter in prehistoric times throughout all the New Mexican and Arizonan 
pueblos. , 

Figure 50, from Chaves pass, was a nicely carved 
imitation of a toad or frog. A somewhat smaller 
shell carving in the form of a frog is figured by 
Holmes in a former report of the Bureau of 
American Ethnology. 

The fragment of a shell which the author is 
unable to identify was found at the Chevlon ruin 
(figure 51). The figure was elongated, with two 
lateral extensions arranged in pairs on each side, 
and suggested a highly conventionalized animal. 
The author has no suggestion to make in regard to its former use, 

and only two specimens of shell carved in this 
shape were found by him. 

Besides these more common shells, many speci- 
mens of Melongena patula, Oliva angulata, and 
Oliva biplicata or hiatula were obtained. 

The crescentic shell ornament shown in figure 52, 
which was evidently hung to some part of the 
body by the hole midway in its length, may have been a gorget, 
or possibly a pendant for a 
necklace. Its form is unique. 

In addition to the specimens ^0ISS^ y '- 

Fig. 50. Shell frog 
from Chevlon (num- 
ber 157833). Length 
about li inches. 

Fig. 51. Shell object 
from Chevlon (num- 
ber 157251). 

Fig. 52. Shell gorget from Chevlon (number 


of sea shells which preserved 
enough of their natural form 
to render identification possi- 
ble, the author collected many 
fragments of unknown relationship. It is probable that the major- 
ity of these belong to some one of the species already mentioned. 
Of unidentified fragments perhaps the most numerous were shell 


beads, of which there were many hundreds. Some of these were 
large and of coarse make, but others were so minute that it remains 
a marvel how they could have been manufactured with the rude 
implements a stone-age people had at its control. In some instances 
the perforations were but a trifle larger than the diameter of a fine 
needle, with rim not over a sixteenth of an inch wide. The thickness 
of these beads was not greater than that of paper. 

All the species of shells which were found in ruins belong to the 
molluscan fauna of the Pacific, and are still used for ceremonial or 
ornamental purposes in modern Hopi pueblos. A majority of these 
have been collected in cliff houses and cavate dwellings, and likewise 
occur in even greater numbers in the ruins along the Gila and Salado 
rivers in southern Arizona. Not a single specimen was found which 
could be traced to the Atlantic watershed, but the source of all was 
the Pacific ocean, or, what is practically for our purposes the same, the 
gulf of California. Still more significant is the fact that the art upon 
them — the symbolism with which they are decorated — is identical 
with that on the pottery of the ancient sedentary people of southern 

It may be said that the simple existence of these shells in the ruins 
from the Gila valley to modern Tusayan can be explained on the the- 
ory of barter, and that their distribution does not prove racial kinship 
of former owners is self-evident. The theory that the same symbolism 
and treatment of the material originated independently can not be 
seriously urged in this case. While there is no proof one way or the 
other that these shells were worked by the people who lived in the 
ancient ruins, it is probable that the ancestors of the Hopis may have 
brought them in their migrations from the south. That the cul- 
ture came to Tusayan from the south appears probable, and Hopi 
traditionists claim that not only their culture, but also the ancestors 
of certain component clans of their people came from that direction 
into Tusayan. So far as archeological researches bearing on this 
problem are concerned, they verify the claim that the remote ancestors 
of the Patki people of Tusayan formerly inhabited the Gila-Salado 
drainage area, and were closely allied to the Pimas, or some other 
tribe of that slope. 

Bone Implements 

The collection of bone implements was Large and varied in character. 
In the specimens from Chaves pass, where, from the nature of the 
country, antelope were abundant, we find a large number of bone 
implements made of the leg bones of the Cervida 1 (see figure 53), 
but in the ruins of the Little Colorado, that is, Homolobi and 
Chevlon, bird bones commonly formed the material from which they 
were made, am! few large bone awls wore found. 



[ETH. ANN. 22 

The only specimen of a carved-bone implement is shown in figure 54. 
It was made from the bone of a deer or antelope, and has a quadruped 
carved on one end, the head and legs being well represented. The 
specimen is one of the few found at the small ruin of the Homolobi 
group near the bank of the Little Colorado, about four miles from 
Win slow. 

The general form of one of the small bone awls is shown in the 

Fig. 53. Bone awl from Chaves pass (number 158097). 

accompanying cut, figure 53. As a rule, the bone of which they 
are made is so worked that the rounded end fits the palm of the 
hand and the sharpened extremity is continued to a needle-like point. 
These awls are made of the bones of different animals, of which the 
turkey is the most popular. They were probably used in ancient 
times in sewing or darning fabrics, possibly in weaving. The modern 

Fig. 54. Carved bone awl from Homolobi (number 157866) 

Hopis used a few years ago in weaving a bone awl not very unlike that 
figured above, but of late they have adopted an iron implement. 

At Chaves pass seven small bone awls were found, at Homolobi 
five, and at Chevlon four. 

Sixteen needle-like bone implements with eyes were found at 
Homolobi; fifteen were obtained at Chaves pass, and three at Chevlon. 

Fig. 55. Bone implement from Chaves pass (number 157867). 

Another bone implement that was common at the Chaves pass ruins 
was shaped like a bodkin. Thirteen specimens were found. 

A number of bone tubes, some of which were perforated and others 
not, were found in the excavations. Some of these tubes had holes in 
the sides, diametrically opposite, and were identified as bird whistles. 
One of these resembles the whistle still used in ceremonials con- 
nected with making medicine in the modern Tusayan rites. 





Bone tube froiri Honiolobi (number 
156898 ) . 

Five specimens of small half-tubes showed evidence of having 
been attached in pairs, as the marking of the binding string on the 
bone is still visible. These were probably whistles, the noise being 
made by ;i thin edge. 

There are several bone tubes which resemble "bone implements'" 
found by Nordenskiold in the cliff houses of the Mesa Verde, where 
they are considered "beads made of the humerus of a large bird, prob- 
ably the turkey. " a Possibly 
the tubes from Chaves pass may 
likewise be beads, which, how- 
ever, is not the case wi h the 
Mesa Verde specimens. In the 
account of the excavations at 
Sikyatki similar bone beads, found about the neck of one of the 
skeletons, are mentioned. 

Sixteen of these bone tube-like objects were secured at Honiolobi, 
three at Chaves pass, and one at Chevlon. 

In addition to the common forms of implements already described, 
man}' other specimens were obtained. Some of these were too frag- 
mentary to make possible 
an identification of their 
former uses. Of these mis- 
cellaneous bone imple- 
ments, eighteen complete 
specimens and several 
fragments were secured 
from the ruins at Honio- 
lobi. A very few objects were found at Chaves pass and Chevlon. 

The specimen shown in figure 57 is one of the most interesting 
bone objects in the collection. It is made from a deer or antelope 
leg bone, is fiat on one side and rounded on the opposite, and tapers 
to a sharp edge at one extremity. It was affirmed by one of the work- 
men employed by the author that this is the shaft of one of the sticks 
used by the stick swallowers in their ceremonies. 

Turtle Carapaces 

Near one of the skeletons in the cemetery at Chevlon were found 
two circular objects made of the carapace of a turtle belonging to the 
genus Chrysemeis, but not the indigenous species. One of these was 
perforated (see figure 58) and the other was not, but when found both 
woe in four fragments — not broken, but having fallen apart at (he 
sutures. The faces were ground smooth and the whole form had 
been considerably changed. These disks may have been spindle 

Fig. 5*. 

Stick used by stick swallower, from Chevlon 
(number 158076). 

oThe Cliff Dwellers of the Mesa Verde. Stockholm, 1898, pi. XL, fig. 22. 



[ETH. ANN. 22 

Horn Objects 

Two fragments of bone or horn for which the author is unable to 
assign any use were dug up at Chaves pass. One of these was per- 
forated, and had a number of indentations on the edge imparting to 
it the appearance of a fire board. It may have been used in kindling 

Two fragments of deer horn were found in one grave. These were 
more or less worn, and one of the Hopis is responsible for the infor- 
mation that chips of similar horns are at the present day drunk with 
water as a medicine. "The deer," he says, "has a good heart." 


The custom of placing a small earthern vessel with different col- 
ored pigments with the dead was practiced by the people of Chaves 

pass, Chevlon ruin, Homolobi, and 
Old Shumopovi. The pigments used 
were the same as those now employed, 
and had apparently the same cere- 
monial significance. As these sub- 
stances are now highly prized, and as 
there is every reason to suppose that 
the3 T were regarded in the same way 
in ancient times, the burial of pig- 
ments with the dead may have been 
of the nature of a sacrifice. 

Some of the modern Shumopovi 
Indians begged for fragments of 
green carbonate of copper which 
were found in the graves of their 
ancestors, for use in painting their 

ceremonial objects and for other purposes. 

Specimens of red paint (sesquioxide of iron), blue paint .(azurite), 

green paint (carbonate of copper), and 

white paint (kaolin) were found at the 

various ruins visited. Some were ground, 

while others were in lumps occasionally in 

the form of a cylinder or disk. 

In his account of the ruins of Sikyatki 

the author called attention to objects in 

the forms of disks, cylinders, and the like, 

which were found in cemeteries of that 

ancient pueblo. Their uses were said to 

be problematic and he now has to record the finding of other objects 

of the same nature and form which are equally enigmatic (figure 59). 

One of these from Homolobi is a hemispherical fragment of kaolin, 

Fig. 58. Disk of turtle shell, from Chevlon 
( number 157841 ) . Diameter nearly 3 

Fig. 59. Kaolin cup from Chaves 
pass (number 157928). 




recalling those exhumed from Sikyatki; there was also a cylindrical 
object of the same material from the Chevlon ruin, but the most 
exceptional specimen was a disk-formed object of kaolin with a 
depression in one side, resembling a small mortar. These various 
forms into which the easily cut kaolin is worked would seem to have 
served some important office, the nature of which is unknown to the 
author. Kaolin at the present day is used for whitening cotton 
blankets, sashes, and kilts, and for painting the bodies and limbs of 
those who participate in sacred festivals. Possibly some of these 
fragments are simply pigments. 


Considering the number of graves oj^ened in the course of the exca- 
vations, it is remarkable that so few specimens of cloth were found. 
This may be ascribed, not to the poverty of the inhabitants of the 
ancient villages in woven fabrics, but to the rapidity with which cloth 
decays in the moist soil. One or two of the specimens which were 
found were preserved by the copper carbonate with which they were 

Fig. 60. Matting from Chevlon (number 157912). About 5 by 2 : ; inches. 

in contact, but the fragments were small and the manner of weaving 
difficull to discover. From one of the specimens it appears that the 
haii- of some animal was used, and there is no doubt from others 1 hat 
yucca fiber was extensively employed. The impression of string was 
observed on several sticks, but the string itself was too much decayed 
for ident ilieat ion. 
Two fragments were discovered a1 Homolobi ami one at Chaves 



It appears that the bodies of the dead, especially at the Chevlon 
ruin, were wrapped in matting, fragments of which were in many 
instances well preserved. This matting was a loosely woven fabric, 
made of vegetable fiber which the author is unable to identify, and 

22 ETH— 04 7 



[ETH. ANN. 22 

was most abundant in the Chevlon ruin. It does not appear to be 
common in the interments at Chaves pass, and was only sparingly 
found at Homolobi. 

In the accompanying cut (figure GO) a portion of one of the best 
specimens of this matting is shown as a type. The specimen from 
which this fragment was taken is large, ample enough to cover the 
head and parts of the shoulders. 


The specimens of basketry found in the ancient burial places were 
of two kinds, one corresponding with that now made at Oraibi and 

Pig. 61. Basketry of Oraibi type, from Chevlon (number 157918) . Length about 5 inches. 

another similar to the basketry of the Middle mesa. It was the cus- 
tom to bury these objects with the dead, as is still a common usage 
in Tusayan, as receptacles for food and as other mortuary offerings. 
These baskets were ornamented with woven patterns, and in some 
instances were covered with a thick layer of green, blue, or red paint. 
In form they were generally plaque-like, but one of the specimens 




which was almost entire was deeper and basket-like. All were 
browned with age, and the majority of specimens were fragmentary. 

One of the best specimens of the Oraibi style of basketry is shown 
in figure 01, which represents a sector of a plaque in 
which the colors are still preserved. The manner of 
plaiting this basket is the same as that practiced at 
the Ilopi pueblo, as is shown by the accompanying 
cut, figure G'2. 

The inhabitants of ancient Chevlon were familiar 
with the method of making basketry by coiling, as 
at the Middle mesa of the Hopis to-day. There are 
many specimens of this ancient coiled basketry in 
the collection, but the specimen shown in figure 63 
is one of the most perfect found. The design is the 
same as that shown on some of the ceramic objects. 

One specimen of coiled ware and two fragments were found at 
Chaves pass, and twelve specimens of coiled ware and two of the 
Oraibi pattern at Chevlon. 

Fig. 62. Basketry of 
Oraibi type, show- 
ing manner of plait- 

Fig. 83. Basketry of roiled type, from Chevlon (number 157915). Diameter about •"> inches. 


Several forms of prayer-sticks, or pahos, were collected from t Indif- 
ferent ruins in L896, bui the majority were so decayed that their orig- 
inal forms were unrecognizable. 

Small prayer-sticks painted green were found at Shumopovi, and 
these were apparently not unlike those now used in tin- neighboring 


The pahos from Chaves pass were painted bright blue, the pigment 
used being azurite. These had a single stick as long as the arm, 
between a half inch and an, inch in diameter, and were apparently 
laid by the side of the bod} 7 when buried, in some instances resting 
on the left arm. So thoroughly decayed, however, was the wood of 
which they were made that it was difficult to take an unbroken speci- 
men from the earth. 

An unusual form of paho was found in a grave at Chaves pass. 
It consisted of a wooden disk with a short handle, and was not unlike 
a hand mirror. The whole object was painted green, with an obscure 
figure in red on one face. The disk was perforated in the middle, 
and there were markings or impressions of feathers on the green pig- 
ment with which it w T as painted. In one of the graves a spheroidal 
wooden object was found, with daubs of pigment upon it, which 
recalls the squash pahos found in the Awatobi shrine, and has like- 
nesses to the modern prayer eagle eggs made of wood, such as can be 
found in a shrine at Tiirkinobi, near Walpi. 

The pahos from Homolobi and Chevlon are small, and as a rule 
are painted green with copper carbonate. They were neither as finely 
made nor as complicated as those of Sikyatki, where the best ancient 
praj T er-sticks yet found have been obtained. In the Chaves pass 
graves the prayer-stick was verj^ long and painted blue with azurite 
or green with carbonate of copper. 

The most interesting and exceptional form of prayer-stick collected 
in 1896 was that obtained at Chevlon. It was disk-like in form, had 
an attachment at one end or on the rim, and was painted green and 
decorated with red designs. The author has found no similar prayer- 
stick in use in any Hopi ceremonial which he has witnessed. 

Bow and Arrows 

Almost a complete bow and several fragments of arrows or reeds 
were dug. out of a grave at the Chevlon ruin. These were taken from 
the same place as the gaming reeds about to be mentioned. The 
association of gaming implements and the weapons of a warrior in the 
same grave is highly suggestive. 

Gaming Canes 

One of the mortuary bowls excavated from the Chevlon cemetery 
contained five half canes (figure 64) which recall those used in the 
Zuiii game known as sholiwe. a These reeds are not, however, marked 
as are those now used at Zuiii, and differ from those found on a 
bowl from Shumopovi (figure 74). The author believes, however, that 
the game in which they were employed was practically the same. 

"See Owens's figure of sholiwe reeds in Popular Science Monthly, May. 1891: According to 
Owens, four pieces of reed are employed in this game by the Zunis. The manner of throwing 
the reeds is clearly indicated by this author. 


One or two food bowls were found in which figures of these gaming 
reeds are painted on the inside of the bowl, as is mentioned later in 

this report. 


Many of the food basins contained seeds of maize or squash, and 

ears' of maize from which the kernels had decayed were found in 

several bowls. 

The ears or cobs show 

that the maize was a 

small variety, like that 

found in cliff houses 

and still cultivated by 

the modern Hopi farm- 
ers. The occurrence of 

squash seeds in some of 

the mortuary bowls is 

important, indicating 

the ancient use of this 

vegetable for food. It 
may, in this connec- 
tion, be borne in mind 
that one of the southern 
clans of the Hopi In- 
dians was called the 
Patun or Squash fam- 
ily, which is still repr e- 
Middle mesa, although 
it is extinct at Walpi. 
Numerous other small 
seeds too shriveled for 
identification were 
found in the mortuary 
offerings in the ceme- 
teries of Homolobi and 

Fig . 64. Gaming canes from Chevlon (number 158080) . Length 

about 3i inches. 

ICIK'S <)i J1UIIHHUU1 mna 

Chevlon. It would seem from the nature of a matrix in which they 
were inclosed thai thev had been boiled or cooked in some way. 


Almost every bowl found in the cemeteries contained fragments of 
whal appeared to have been food, but in most instances t his was too 
mud, destroyed to be identified. It was ordinarily in the form of a 
thin film coating the interior of the bowl, and was penetrated by 
roots which had found their way from the surface of the ground. 
There Ls Little doubt thai in some instances this food was one of the 
many kinds of corn bread so common anion.-; the modern Ilopis. 



[ETH. ANN. 22 

Stone Implements , 

In the type "pounding stones" are included stone implements, ordi- 
narily of small and medium sizes, showing marks of battering at one 
or both cuds. Apparently they were not attached to a handle, and 
the indication is that they were simply held in the hand when in use. 
With these pounders various substances were bruised, pigments 
were ground, hide was made more pliable, and similar processes 

were effected. Any 
rounded stone conven- 
ient to handle appears 
to have been taken for 
this purpose. At Homo- 
lobi thirteen stone ob- 
jects were found which 
were doubtless used as 
pounding implements, 
at Chaves pass four, and 
at Chevlon two. 

Stone axes were even 
more common. At Ho- 
molobi twenty-four specimens were found, at Chevlon two. 

Several weapons made of a black stone, one of which is shown in 
figure 65, were collected 'at Homolobi. These were evidently either 
celts or spearheads, for there was no sign of hafting or of polishing, 
and marks of fracture were apparent. 

The specimen represented in figure 66 is of white stone, possibly 
aragonite; it was the only weapon of .this material which was found. 
The shape is regular and the surface smooth, and there is a groove for 
a handle. It has four grooves cut on one edge and incised crosses, 

Fig. 65. Stone implement from Homolobi (number 157895). 
Length about Gi inches. 

Pig. 66. Stone ax from Homolobi (number 157024). Length about 7 inches. 

two in number, one of which is shown in the cut. No indication of a 
handle accompanied this implement, which leads the author to sup- 
pose that that part was of wood, which had long ago decayed. 

There were several stone implements with a cutting edge on one 
side which were probably used as knives. These are sometimes 
curved, bu1 generally straight. The best stone knives are two from 
the Chevlon ruin. 


Several of these knife-like objects had dentate edges, as though 
used for saws. 

The author has elsewhere noted the frequency with which fragments 
of obsidian, and chips from the same, occur in ancient Pueblo graves, 
or on the surface of mounds, indicating ruins. The three sites 
of explorations in 1896 were not exceptional in this particular, and a 
somewhat limited amount of material of this nature was collected at 
Homolobi, Chevlon, and Chaves pass. This material was prized in 
ancient times for arrow points, spearheads, and knives, and the sharp 
edges of many of the chips were probably used for cutting. 

Arrow points were found in large numbers at the various ruins 
visited on the Little Colorado. 

The present Hopi Indians use a grooved stone for polishing arrows. 
These stones are ordinarily of a coarse sandstone, which acts as a file 
on the wooden object rubbed back and forth in the groove. Similar 
polishing stones are very common in ruins, assuming a number of 
shapes, and made of several different kinds of rock — as sandstone, 
lava, slate, and even clay. It is probable that these stones were used 
for the same purpose as those employed in the modern kivas, as their 
form has not changed from ancient times. 

The differences in the polishing stones from the ruins at Homolobi, 
Chevlon, and Chaves pass were not very great, as the simplicity of 
the implement admits of but few varieties. Those from Chaves pass 
were made of lava, which occurs only rarely in the other two ruins. 
There were double- as well as single-grooved polishing stones, and 
shallow- and deep-grooved ones. It is possible that some of the per- 
forated stones may have been used as polishers for arrows in much 
the same way as the modern grooved stones are employed. One of 
these anow polishers had the shape of an animal, and was narrowed 
to a head at one end. On this end was cut a mouth and two depres- 
sions in the proper positions for eyes. The surface of the stone on 
the side opposite that occupied by the groove was flat and smooth, so 
that tin* object could be firmly placed when in use. 

It appears to have been a mortuary custom among the people who 
lived in the ancient pueblos along the Little Colorado to deposit with 
women and girls at death a metate and its grinder. These were ordi- 
narily reversed when buried, as though symbolic of the death of the 
one who formerly used them. In one of the Homolobi graves three of 
these metates with their corresponding hand stones were found, and 
these were added to the collection. They were made of rocks of dif- 
ferent degrees of smoothness, and were evidently formerly used in 
grinding coi'D in 1 he same way as in modern Hopi pueblos. The maize 
was bruised and roughly ground on one of these stones, then it was 
pilled to a finer-grained one, and ultimately to the finest of all. These 
metates were much worn, showing long and constant U? 


.Many large and very heavy metates made of lava rock were fonnd 
at Chaves pass, but these were not brought to the railroad. Several 
of these were worn so deep in one face as to form troughs. If there 
were no other evidence of long occupation of this ruin the deeply 
worn metates would furnish it. These metates were bulky, and were 
quarried in the bad lands of the vicinity of the old pueblo. It is 
instructive to note that metates, rather than smaller and more port- 
able stone objects, show the influence of geological environment, for 
as a rule their size requires that they be made of the rock contiguous 
to the ruin. Sandstone metates are commonest in the valley ruins, 
lava in those built of lava rocks. The same is true of the stones of 
which the walls of the ancient habitations were constructed. 

The material of the grinding stones is the same as that of the 
metates. They are for the most part simple elongated slabs, some- 
times with depressions along the sides to enable the manipulator to 
obtain a stronger hold. Like metates they show the effect of wear on 
one face, being generally rounded on the upper surface. Several of 
these stones are double faced, or when seen in profile they are trian- 
gular. These specimens, which are of softer and finer rock than the 
others, were undoubtedly used in the last stage of grinding the 
kernels of corn into fine meal. Ten of these meal grinders were found 
at Homolobi and Chevlon. 

Stone implements of mushroom form are among the interesting 
objects obtained in our excavations. These are well adapted for 
paint or pigment grinders. 

The mortars in which paint is ground are ordinarily small stone 
slabs with a depression in one face. These are sometimes rectan- 
gular in form, often circular, and the depth of the depression varies, 
being frequently very slight. The same variety of paint mortars is in 
use to-day, and many of these stone objects now used in the kivasare 
no doubt very ancient. The small stones with one flattened face 
were doubtless used as grinders; in some instances they are much 
worn. The list of paint mortars includes two from Homolobi and 
three from Chevlon. 

Stone Slabs 

The use of slats or slabs of stone decorated with figures painted in 
various pigments has been described in accounts of several Hopi 
altars. One of the most remarkable of these is probably the stone 
called the Hokona mana, or Butterfly virgin, which is used in the 
construction of the Antelope altar at Walpi. Two flat rectangular 
stones stand back of the sand picture of the Antelope priesthood in 
the Snake rites at Shumopovi, while the use of similar stones in the 
various Flute altars may be seen by consulting articles on these rites. 

Of a similar nature, no doubt, is the painted stone slab shown in 
plate xlvi, found in a grave at the Chevlon ruin. This object, which 
is much larger than any of those which have been mentioned, is painted 


on both sides with highly suggestive designs of a symbolic nature. 
The decoration on one side is almost wholly obliterated, but on one 
corner we detect clearly the modern symbols of the dragon fly. The 
pigments with which this stone is painted were easily washed off, 
and this accounts for the loss of the decoration on the surface 
which was uppermost as it lay in the grave over the body. The 
design on the other face, however, is more distinct. It consists 
of three triangular figures inclosed in a border, recalling a sand 
mosaic such as is used in modern presentations of the Hopi ritual. 
Two colors, black and white, are readily detected in the border — 
the black outside the white. The field inclosed by this border is yel- 
low, and the three triangular figures are black, with inclosed rec- 
tangles, which are white. At the apex of each triangle there is a rude 
figure of a bird painted red, in which the head, body, and two tail 
feathers are well differentiated. 

The whole character of the design on this stone calls to mind the 
decorations on the walls of a kiva of a cliff dwelling of the Mesa 
Verde, described by Nordenskiold, and figured in his beautiful 
memoir. In the designs on the kiva wall of "ruin 9" we find groups 
of three triangles arranged around the whole*estufa at intervals 
on the upper margin of a dado, and each of these triangles is sur- 
rounded by a row of dots. The field on which they are painted is 
yellow, and the triangles and dots are red or reddish brown. On a wall 
of Spruce Tree house Xordenskiold found a similar dado with tri- 
angular designs, and it is interesting to note that in the figure of this 
ornamentation which he gives rude drawings of birds appear in 
close proximity to the triangles. 

The interpretation of these figures must be more or less hypothet- 
ical. The custom of ornamenting house walls with a series of trian- 
gles on the upper margin of a dado is still observed in the modern 
Hopi villages, where, however, the position of the triangular designs 
is reversed as compared with that of those on this stone slab. 

The triangle is a symbol of the moth or butterfly, which, while 
appropriate on women's blankets or house walls, would hardly appear 
to have special significance on the slab in question. Still, as has been 
pointed out, oik' of the most venerated objects on the Antelope altar 
lias the figure of a butterfly upon it. 

Much more likely is it that these three triangular figures sur- 
mounted by birds are rain-cloud symbols, and that, 11ns slab of 
stone was formerly used in a ceremonial which had for its object 
rain making, and to this conclusion the dragon-fly symbols on Ihe 
reverse side also point. This sione is an altar slab with rain-cloud 


In the Chevlon ruin the author found several flat stones, one of 
considerable size, which were marked with blackened circles. The 
largest of these, fully 3 feel square, was not brought to Wash- 
ington, and the photograph which was made of it soon after it was 


removed from the grave was a failure. These circles, apparently 
made by smoke, are of unknown origin and use. On the largest stone 
they are arranged in two rows, four in each row, the peripheries touch- 
ing. The stones were inverted when found, and occurred in the cem- 
etery at Chevlon only. 

Many graves at Chevlon and Homolobi were, as has been stated, 
indicated by upright stones or flat slabs of rock. As the digging went 
below the surface it was sometimes found that the skeleton was covered 
bv a similar flat rock, and in a limited number of cases these rocks 
were perforated. The holes were sometimes not larger than a broom 
handle, often capacious enough to permit the insertion of the arm, and 
in one instance a foot or more across. Oval, round, and rectangular 
orifices were found, and in several cases a considerable amount of 
labor must have been expended in making them. 

Slabs with the smaller circular holes were also found in the floor of 
a room where there was an intermural burial. Explanations more or 
less fanciful have been suggested for these perforated stones, one of 
which was that the rock had been placed above the body and the hole 
in it was for the escape of the soul or breath-body. The slabs were 
found above the bodies of several deceased Homolobeans, and the 
modern Hopi interpretation of the perforation is offered for what it is 


Small disks were found in all the ruins which were studied, and 
while these had like forms they were made of various substances, as 
of stone, pottery, and shell. They are generally circular in form, rarely 
perforated, and often ground on their edges. The unperf orated speci- 
mens are supposed to have been formerty used to cover a hole in a jar 
in much the same w T ay as similar fragments are now used in flower-pots. 
The perforated specimens were probably used in much the same way 
as wooden disks are employed in modern pueblos, as parts of drills for 
perforating stones, shells, or other hard substances. The specimens 
of this problematic group of objects, and the localities from which 
they were gathered, are mentioned in the appended list. 

Number Locality and material 

158056 Chaves pass; stone 

157706 Chaves pass; red pottery 

158079 Chaves pass; red pottery 
158078 Chevlon; red pottery 

158080 Chevlon; red pottery 
157963 Chevlon; stone 
156480 Homolobi; pottery 
157965 Kisakobi; pottery 
158093 Little Colorado ruins . 

158164 Jeditoh; shell 

158165 Jeditoh; shell 
158060 Chaves pass; galena 
158095 Chevlon; wood 



The number of graves in which smooth, waterworn stones, quartz 
crystals, and fossil cephalopods occur in other ruins has been noted 
in the account of Awatobi and Sikyatki. A considerable number of 
these objects were found in the Little Colorado ruins; the fossils 
included a tooth and remains of several crinoids and of an ammonite. 

Similar objects arc still used in Hopi ceremonies, and it is well to 
call attention to the fact that some of the priests begged the author 
to give them these ancient objects that they might use them in the 
preparation of medicine and in other sacred or ceremonial ways. 

Some of the smooth stones may have been used in polishing 
pottery, but this can hardly be said of the quartz crystals and the 
bot ryoidal specimens. 

Fossil cephalopods, called koaitcoko, although very common in the 
rock strata underlying the modern pueblos, are sometimes looked 
upon by the Ilopis with great reverence, an.d are used in several mod- 
ern ceremonies. One of the best- known instances is in the tiponi of 
the Lalakonti, described in an account of the unwrapping of that 
palladium, as follows: 

The chief i)riestesses and Kwatcakwa then untied the bundles upon the altar. 
They first unwrapped the buckskin thong which bound one of them and took from 
the top a large number of sticks of different lengths, to each of which numerous 
feathers were tied. In the midst of these sticks there was an ear of popcorn sur- 
rounded by a mat made of eight black feather-sticks tied together. This mat, 
surrounding the corn, rested upon a cloth. Removing this cloth from the cradle, 
there appeared below it a nicely folded piece of buckskin painted on the border, 
with the rain-cloud ornaments painted black, the falling rain being represented 
by fringes. Within this skin there were many breath- feathers and a single reed, & 
to which feathers were tied. Below the buckskin there w T ere many bean and 
melon seeds. Within the bundle of breath-feathers there was a fossil shell, frag- 
ments of another, and the pinon branch. The basket itself, which forms the 
cradle was made of a continuous coil of wickerwork, rectangular in shape. 

When the priestess had undone this bundle, the contents of which were consid- 
ered so sacred that we were not allowed to touch them, she carefully repacked it. 
She first put in the seeds of beans, corn, and melons, and then a number of breath- 
feathers. She sprinkled these with metallic iron dust [micaceous hematite] and 
added a pinon branch (]>ine needle). On these she placed the cloth in which the 
fossil shell was tied and the reed with its feathers. Above this she tied around 
the ear of corn the old prayer-sticks, to which she added a new one which she 
had prepared. The black sticks were said to be old men and the seeds to be food. 
The fossils, which are called koaitcoko/' were found later to be one of the numer- 
ous cephalopod fossils abundant in certain places. It was said that these sacred 
specimens came up from the under world. The contents of each bundle were sub- 
stantially the sain*-. 

Ii will be seen from this quotation from a description written in 1802 
thai one of the most sacred objects iii the bundles before the reredos 

a The Lalakofiti; A Tusayan Ceremony, in the American Anthropologist, v. 5, i». 121, April, 1892. 
'' T 1 j i -- resembled the so-called reed cigarettes used in other ceremonies. 
c The same name was giv. m for the whole bundle, 


of the Lalakonti altar was a fossil cephalopod, which, so far as could 
be judged, was of the same species as some of those taken from the 
Little Colorado ruins in 1896 and from Sikyatki in 1895. 

The .ceremonial use of fetish stones in modern Hopi rites has been 
described by the author in several publications, from which the fol- 
lowing quotation a is taken as one of the most complete : 

Saliko brought from her house six ears of corn, a crenellate vessel [medicine 
bowl] . and another bag of fetishes. * * * Saliko took a handful of meal from 
a tray at the poiiya [altar] , prayed upon it, and then, kneeling about 4 feet in 
front of the altar, sprinkled intersecting lines. She placed the crenellate vessel 
in the center, and then arranged ears of corn upon the lines, beginning at the 
northwest, where she placed a yellow ear, followed by a blue, red, white, black, 
and an ear of sweet corn, as shown in the diagram. From her bag she took out 
six smooth waterworn pebbles, the largest of which was H inches by three- 
fourths of an inch in size, and placed them close beside the ears of corn. Begin- 
ning at the yellow ear. she laid down by it a piece of opaque quartz with a 
smoky iron streak; at the blue, a piece of the same with a faint bluish tinge; at 
the red, a piece with a reddish tinge; at the white, a piece of translucent quartz; 
at the black, a piece of shining black iron ore, and at the last, a crystal of bluish 
quartz. ( 

Less detailed is the author's description of the use of these rock 
crystals in another Hopi ceremony : b 

The priest, Ametola, first made a bed of fine field sand on the floor, and 
then rapidly traced on the sand three cross-lines of meal, corresponding to the 
six primary directions. Over their junction he placed a medicine bowl, but not 
that before the altar. Around the bowl he laid, at the ends of the lines of meal, six 
ears of corn, with points directed toward the bowl. Beside each ear of corn he 
placed an aspergill and a rock crystal. Within the bowl he dropped several rock 
crystals and a little honey. 

In the Naacnaiya, or New-fire ceremony, smooth pebbles and 
quartz crystals likewise play important parts in making medicine: 

He placed the first group of six skins upon the meal lines, as indicated in the 
diagram. He then arranged the ears of corn upon the skins, and close beside 
them he placed the six pebbles (each having some requisite peculiarity, but no 
opportunity offered to examine them closely) , and finally another set of six skins 
was deposited upon the right of those first laid down. . . . 

Eight songs were sung while he was placing these objects, and during the sing- 
ing of another group of eight songs the asperser laid the pebbles in the nakwipi 
[medicine bowl] , and then rested the ears on end within it. He then slightly 
dipped the tail or the distinctively colored end of each bird skin and each feather tip 
into the water, afterward laying it down in the place from which he had taken it. 
He also sprinkled pollen in the bowl, and aspersed to the six directions with each 
ear of corn as he took it out and laid it in its former place. The song was an hour 
and a half long, and just as it closed the asijerser took from one of the chief's bags 
a quartz crystal. Sucking it, he passed it to a young man sitting near, stitching 
a kilt, who went ujj the ladder and reflected a ray of sunlight into the nakwipi, 
and afterward the crystal was put into the liquid. c 

In the Niman Katcina (festival celebrating the departure of the 
katcinas) we find these crystals and medicine stones likewise used 

a American Anthropologist, v. 5, p. 221, July, 1892. 
b American Anthropologist, v. 5, p. 117, April, 1892. 
c Journal of American Folklore, v. 5, p. 192-193. 


in raising the charm liquid, as the author has elsewhere described as 

follows : a 

Upon the ears of corn were then laid the nakwiowa. small quartz crystals, 
elongated, black, pipestem bodies and shell beads. One or two of these were 
placed on the end of each ear of corn at a point nearest the nakwipi, one behind 
the other. Two similar quartz crystals were placed in the nakwipi. These 
sacred stones were laid on the corn in the same order as above mentioned for the 
ears of corn and aspergills. Intiwa's assistant then took a quartz crystal, one of 
the nakwiowa. and, standing on the lower round the ladder leading out of the 
kiva . held it in the direct rays of the sun and threw the reflection into the water 
contained in the nakwipi. This ceremony, by whijh u ray of light from the sun 
(Dawa) is introduced, was observed in silence. . . . After several strains of 
this and other similar songs had been sung, Intiwa's associate took up the ear of 
corn on the north side of the bowl, held its end over the nakwipi, and, pouring 
water (liquid) from a small spherical gourd upon it, washed off the medicine 
stones into the nakwipi with great solemnity. . . . After the washing of 
these stones into the liquid, a song with quicker time and more rapid motion of 
the rattles was taken up and continued for some minutes. 

The above quotations from descriptions of ceremonies, and others 
which might be mentioned, show how often small waterworn pebbles, 
or quartz crystals, are used in making charm liquid or "medicine" 
by the modern Hopi Indians, and the same kind of stones discovered 
in ancient graves probably had a similar use among the ancients. 
Indeed, one of the quartz crystals from a grave was appropriated for 
his altar by a Shumopovi priest. 

The use of two small stone cylinders, one of lava (number 157984) 
and the other of a light gray stone (number 157983) each having a pit 
or depression in the end, is problematic. Similar shaped stones are 
sometimes used in modern Hopi ceremonies to indicate the zenith. 

Among the fetishes found in Homolobi graves may be mentioned 
a rude bird made of unburnt clay, 6 the only specimen of animal effigy 
which was found in all the excavations in 189G. On the surface of 
one of the mounds of the smaller ruin at Chaves pass, however, the 
author picked up a small imitation of some unknown animal, which 
was rudely carved, and reminded him of the so-called "hunting 
stones," or fetishes, of the Zunis. 

The occurrence of rude effigies of animals in prehistoric Arizonian 
graves may possibly be interpreted as substitutional sacrifices, and if 
this interpretation is correct, it would seem thai in ancient times 
birds, which are now represented by effigies, were sacrificed. It is 
much more Likely, however, thai these images represent animals which 
the ancient worshipers desired, and thai they were forms of prayer 
by signal ares. Small figurines of domesl ic animals are made for 1 his 
purpose at the present day by the Hopi priests. 

a Journal of American Ethnology and Archaeology, v. :.'. L892, p. 75, 76, 77. 

bThis effigy is very unlike the burnt clay imitations <>f birds which have been described in the 
author's account of the ruin A watobi. It is very rudely made, simply pressed into shaped by the 

firitrt-7--. and is without decoration. The Awatobl clay birds are probably used as pendanta, 
while that found at Homolobi appears i" be simply a mortuary offering. 


Human Crania 

A particularly fine collection of crania was obtained in 1896 from 
Chevlon and Chaves pass. One specimen from the latter place had 
the facial and frontal bones stained green. 

Animal Remains 

Although the prehistoric Pueblos of Arizona and New Mexico appar- 
ently had no sheep, horses, or cattle, they domesticated several ani- 
mals, and used many more for food. The bones of these animals 
occur both in houses and graves — more abundantly in the former, 
especially in the Homolobi ruins, where the author made a consider- 
able collection of them. The} 7 have been identified by Mr F. A. 
Lucas, of the National Museum, and are as follows: 

Canis familiaris, domesticated dog; a cranium of the broad-skulled 

Eskimo type. 
Canis frustror, coyote; probably jaw and leg bones. 
Felis concolor, puma; jaw of young. ( 

Lynx rufus, wild cat; jaw. 
Taxidea berlandieri, badger; two claws. 

Cariacus macrotis, mule deer; leg bones, antlers, and part of jaw. 
Antilocapra americanus, antelope; leg bones, part of jaw, and horns. 
Castor canadensis, beaver; collar bone. 
Lepus texianus griseus, jack rabbit (many specimens). 
Lepus arizonae minor, small rabbit (many specimens). 
Cynomys, prairie dog; skull and odd bones. 
Ovis canadensis, mountain sheep. 
Spermophilus, ground squirrel: tooth. 
Thomomys, gopher; skull. 
Corvus corax sinuatus, raven; wing bone. 
Bubo virginianus, great horned owl; tarsus. 
Aquila chrysaetos, golden eagle; claw. 
Buteo borealis, red-tailed hawk; wing bones. 
Cathartes aura, turkey buzzard; wing bones. 

Meleagris gallopavo mexicana. turkey (many specimens — some of young). 
Grus mexicanus, sandhill crane; wing bones. 
Anas sp., duck; part of sternum. 
Phalacrocorax sp., cormorant; 1 bone. 

Miscellaneous Objects 

The manner of building the roofs and floors in the Homolobi ruins 
was practically the same as in the modern Tusayan pueblos. A large 
number of rafters were taken out of the walls, many of which were in 
place, while some had fallen in, broken by superimposed weight. In 
several instances these beams were well preserved; in others they 
were much decayed. Several fragments of the clay with which the 
roofs were covered were collected, and in one impressions of reeds 
were evident. 

Asbestos appears to have been considerably prized by the inhabi- 


tants of tlie Chaves pass ruins, and a single specimen was added to the 
author's collection from these ancient towns. 

A few fragments of a bird's egg, too broken to be accurately iden- 
tified, were collected in a grave at Homolobi. This was possibly an 
eagle's egg, and it may be mentioned that in certain Hopi ceremonials 
at the present day imitations of eagles' eggs made of wood are at 
times placed in modern shrines. The author has no knowledge, how- 
ever, of a modern mortuary use of birds' eggs, but suspects that the 
egg of the turkey, which we know was domesticated by the ancient 
pueblo people, may have been eaten by them. If this supposition be 
well founded, the fragment of birds' eggs in a grave at Homolobi may 
be a remnant of food offerings. 

A single specimen of galena was taken from a grave at Chaves pass. 
This mineral was probably used as a pigment, but it is not common, 
and is not used by the modern Hopis in painting the bod} 7 or any of 
their ceremonial paraphernalia. 

A small copper bell was found in a grave at Chaves pass. This 
object is shown in figure 67, from two faces. It was apparently 
formed of beaten native copper, but the eyelet would seem to indicate 
a knowledge of soldering. One side was so broken that a small spher- 
ical body which served as a clapper could be 
easily seen. The metal was much corroded, 
but not so much as to prevent the bell emit- 
ting a sound when shaken. 

Among other miscellaneous objects taken 
by the author were a fragment of potter's clay fig. 67. Copper beii from 
from Homolobi, a sulphur nodule and several about finch* iame 
dried lizard tails from Chaves pass, a frag- 
ment of asphalt and some perforated cedar berries from Chevlon. 

Several rectangular fragments of red pipe clay, one of which was 
perforated as though for suspension, were found at Homolobi. 

The object of the large and small rectangular or trapezoidal plates 
of mica and selenite found at Homolobi and at Chevlon, in graves and 
elsewhere, was not wholly clear. No specimen, however, was brought 
back from Chaves pass, and no natural deposits of selenite were 
noticed in the latter locality. 


( r kneral Features 

Although in the report of the author's excavations at Awatobi and 
Si U \ at ki, in 1895, an extended account lias been given of the arche- 
ology of pueblos near the East mesa, no work was done on the numer- 
ous ruins at Oraibi and the Middle mesa. The author was particu- 
larly anxious to compare pottery from some of these ruins with the 
beautiful series which had been collected in 1895 at Sikyatki, espe- 


cially as numerous specimens had been sold to traders from Old 
Shumopovi, and almost all of these were identical with those from 
the East mesa. 

There is evidence that Shumopovi was one of the oldest settlements 
on the Middle mesa, but the legends of this pueblo have never been 
carefully studied, and the component clans are practically unknown. 
The pueblo stood in the foothills near a spring when the Spaniards 
first came into the country, and its name can easily be recognized in 
Espejo's list of Hopi towns at the end of the sixteenth century. . 

There is a uniformity in statements that the founder of Oraibi, 
Matcito, lived in Shumopovi before he sought the cave where he lived 
when Oraibi was built, and probably Shipaulovi was founded by 
clans from it in the eighteenth century. The size of the old ruin 
shows that in ancient times it had a large population. 

The record of work at Old Shumopovi can be given in a few 
lines. For several years it has been known that a wealth of beautiful 
pottery lies buried in the cemeteries of that ruin (plate lii). That 
the number of objects destined to be removed from this place is large 
is probable from the great size of the cemeteries and the small portion 
of them which has vet been dug over. The author therefore beiran 
work with high hopes of a great harvest. About twenty Indian 
workmen from the East mesa and a few from Shumopovi were 
employed as excavators, and in the first two days of work they took 
out of the burial places over one hundred specimens. 

On the evening of the second day the chief of Shumopovi, Nacihip- 
tewa, went to the camp of the workmen and forbade them to continue 
the work. It seems that the chiefs of the three villages, Mishongnovi, 
Shipaulovi, and Shumopovi, had assembled in council on the night of 
the author's arrival and decided that his work should not go on. 
For some reason they had not communicated their wishes to the 
author, but went directly to the Indians, working on their feelings 
and threatening them with trouble if they continued excavating. As 
soon as the author learned of the objection he immediately called a 
council of the chiefs at his camp, and learned from Nacihiptiwa that 
he did not wish the work to go on, fearing that it would cause great 
winds which would drive away the rain clouds. The author respected 
his wishes and ceased work at Shumopovi, discharging his workmen. 
Had he been able to complete the work at this ruin there is little 
doubt that over a thousand bowls could have been taken from the 
burial places of that ancient pueblo. 

It would appear from the examination of the Shumopovi cemeteries 
that they were distributed among the foothills east of the main spring 
of the present town. From the quantities of broken pottery in this 
region, it is evident that their extent was very considerable. In some 
instances burial places were separated a considerable distance from 
the ruins of the pueblo, in others they were quite near the founda- 
tion of the Walls. 



A small cemetery was discovered about a quarter of a mile east of 
the ruins, whore there is a patch of sand in which grow a few dwarf 
peach trees. The author camped at this point, which was as near as 
he could approach the ruin with his wagon, and immediately after his 
arrival a family of Shumopovi people came down from the mesa and 
began to hoe the squash plants which grew there — an act which was 
interpreted to mean possession. The Snake chief of Shumopovi had 
a brush house, called a kisi, overlooking his farm, on a small hillock 
near this burial place. 

The ancient pueblo can be traced for several hundred feet, but its 
old walls have been buried or leveled, and very few evidences of its 
architectural plan can be made out by superficial studies. The 
mounds of the old site are covered with fragments of pottery of the 
finest character, beautifully ornamented," with the characteristic 
Sikyatki symbol 

Pottery from the Ruins 

general features and form 

A superficial examination of the pottery of this old pueblo shows 
what a more intimate study demonstrates —that it is very similar to 
that from ruins near the East mesa, and that it differs from that of 
the Little Colorado pueblos. The majority of the pieces belong to the 
fine yellow ware (plates xlviii, l), smoothly polished and elaborately 
decorated. There are a few examples of red and black ware and 
one or two specimens of black and white ware (plate xlix), but the 
yellow ware predominates, as it does at Sikyatki. This is undoubt- 
edly due to the chemical constituents of the clay used in its manufac- 
ture. There are no specimens of red, black, and white ware, and no 
black and no glazed varieties. 

The resemblance, which amounts almost to an identity, in the char- 
acter of the pottery of Sikyatki and Old Shumopovi, as well as 
the similarity in the symbolism, adds weight to the belief that these 
pueblos were inhabited synchronously. 

There is no essential difference in the shapes of the pottery from 
this ruin and of that from the pueblos on the Little Colorado, where 
food bowls, vases, jars, ladles, and dippers are represented. In plate 
li two of the best specimens are figured, but there are many others 
in the collection of equal beauty which have not been reproduced. 

There were two specimens of food bowls from Shumopovi with 
snouts, one of which is shown in the accompanying cut (figure (>8). 
This form appears to be rare, and has not yet been found in the 
ruins alone the Little Colorado river. 

u A very fine collection of morl uary pottery was made at Shumopovi by Mr T. V. Keam, after 
the author was obliged to abandon work there. Many of these pieces are now in the Field 
Columbian Museum at Chicago, which purchased the collection. 

22 eth— 04 8 



[ETH. ANN. 22 


The picture writing on Shnmopovi ware closely resembles that on 
the ware of the ruins near the East mesa, and though from the limited 
examination which was possible on account of the size of the collection 

Fig. 68. Bowl with snout, from Shumopovi (number 157817). 

few new forms were found, the author is able to add some instructive 
pictures to those already known from this region. On the whole, Old 
Shumopovi pictography is like that of Sikyatki and Awatobi, and 
differs from that of Homolobi. The geometrical figures do not widely 

Fig. 69. Plumed Snake. Design on food bowl from Shumopovi (number 157769). 

differ from those of other pueblo regions in Arizona north of the 
Mogollones, and are of the same type as those of Chaves pass, Chevlon, 
and Homolobi. 




Plate LI shows the face of a masked dancer, the treatment of the 
left eye of which resembles that of this organ in certain Zuni masks. 

The only figure of a reptile which was found at Shumopovi was 
drawn on the inside of a food bowl (see figure 60). This figure is so 
different from any representation of the Plumed Snake that the author 
has hesitated to refer it to this mythic being. The feathers on the 
head, if such they be, are two semicircular bodies, and the tongue is 
represented by a line with arrowpoint termination. The eyes are both 
on one side of the head, and the lines on the head and body are incised, 
making designs which are highly successful from a decorative point 
of view. The bowl is a small one, and is made of the fine yellow ware 
characteristic of ancient Tusayan ruins. 

An examination of bird figures from Shumopovi shows a marked 

Fig. 71 1. Mythic bird and game of chance. Design on food bowl from Shuniopovi (nnmber 157714). 

difference from those of the ruins on the Little Colorado and a close 
likeness to those of Sikyatki and other ruins near the Easl mesa. 
Specimens were found with the peculiar conventionalized form of the 
"breath feather" so constant in the collections made in L895, and there 
were fine specimens of the sky band and the dependenl bird. The 
design represented on the food bowl shown in figure 70 is wvy instruc- 
tive. From a comparison with other figures of Kwataka the author 
is led to refer this figure to the mythic bird-man god. The head rep- 
resented in profile lias two triangular feathers, and on the throat and 
breast appear the terraced designs so often found in bird symbols. 
The feathers of the wing are triangular. There are reasonsof a com- 
parative nature which lead the author to believe that the band on 
which the bird rests represents the sky band, and the ring represents 



[ETH. ANN. 22 

either the earth or the sun. The position of the three tail feathers 
in this drawing is thoroughly characteristic of ancient pueblo art. 
Instead of being drawn in a horizontal plane, as they naturally would 
be in a side view of the bird, they appear in a vertical plane, as often 
occurs in these figures. This characteristic arrangement of the tail 
feathers is common in the decoration of modern vases from Acoma, 
where the bird is a constant ornament. It may also be seen in the 
avian figures which decorate many of the ancient bowls, vases, and 
jars from Sikyatki and Awatobi. The terraced figures on the under 
side of the head are of frequent occurrence in bird designs. The three 
cross lines occur on several Sikyatki bowls and represent stars. 

The design on the food bowl from the ruin of old Shumopovi, shown 
in figure 71, represents a bird god, as is shown by the three tail feathers 

and the triangular wings. The head 
takes the form and bears the symbol- 
ism of that of a masked katcina still 
personated in Tusayan. The horn on 
the left side of the head terminates 
in a conventional figure of a feather, 
and the design on the right-hand side 
is a symbol of the squash flower. On 
the face is the terraced symbol of a 
rain cloud, still used in modern Hopi 
symbolism and very common in an- 
cient bird figures. Within this ter- 
raced figure are represented the dragon 
fly, rainbow, and falling rain. It will 
be noticed that each of the two exte- 
rior tail feathers bears two smaller 
Avhite lines. Similar symbols charac- 
terize the figures of the war god, and 
are said to indicate the hawk. There 
are legends extant that these are 
markings made by the claws of some 
animal in its struggles. They are 
found on the cheeks Of idols of the war god in several pueblos, as Sia, 
Zuni, and those of Tusayan. 

The avian figure on the food bowl shown in figure 72 represents 
a raptorial bird with extended wing. The homology of the two long 
bodies dependent from the breast is apparent when we compare them 
with the symbolic feather on Sikyatki pottery. They represent the 
breast feathers of the eagle; the symbol is still preserved in modern 
Hopi ceramics, but, so far as is known, has not yet been found on 
pottery from the Little Colorado ruins. 

Both eyes are represented on one side of the head, and the beak is 

Fig. 71: Bird design on food bowl from 
Shumopovi (number 157795) . 




curved like that of raptorial birds, which are so common in the deco- 
ration of Ilomolobi ware. This figure also shows a very common vio- 

Ftg. 72. Mythic bird design on food bowl from Shumopovi (number 157134). 

lation of perspective among ancient and modern Hopi artists, for the 

tail feathers are turned from the natural horizontal to a vertical plane. 

The design shown in figure 73 represents a bird in which the curved 

Fig. 73. Symbolic bird design on food bowl from Shumopovi (number L57771 ). 

body above is the head, t lie two Lateral triangles the wings, and the 
three pointed bodies the three tail feathers. The two curved bodies, 



[ETH. ANN. 22 

one on each side of the tail, are backward extensions of the bodies 
which assume different forms in as many different representations. 

This is a form of bird symbolism unknown in pictography from the 
Little Colorado ruins, but very common, in many variations, at Sik- 
yatki. Its reduction to bird symbolism maybe readily followed by a 
comparison with the series given in the report on the 'field work of 


'. V 



Fig. 74. Gambling canes and bird. Design on food bowl from Shumopovi (number 157735). 

The food bowl shown in figure 74 is a beautiful specimen of yellow 
ware, decorated on the interior with two figures, one representing a 
bird and the other four canes used in a game still played in modern 
pueblos. The bird figure evidently represents the Heart-of-the-sky 
god, whose symbol is a star, which is represented on the head of this 
divinity in designs from Sikyatki figured in the account of the expe- 
dition of 1895. The wings and tail feathers, three in number, are 
easily recognized. 

The four gaining canes are marked in different ways, and corre- 
spond with the four cardinal points. Their markings are, however, 

"Seventeenth Annual Report of the Bureau of American Ethnology, pt. 2, 1898. 


different from those on a set of these canes taken from the rain at 
the month of the Chevlon fork, to which reference has already been 

These canes, corresponding in a general way with those used in 
Zuni in the game toshalewa, are rarely employed at the East mesa, 
but the occurrence of figures of them in old Tusayan ruins indicates 
the age of this game in the pueblo area. While the markings on these 
objects are not the same as those on the Zuni, the variations are no 
greater than would be expected, considering the ruin in which they 
were found. The general character of the game was evidently very 


In continuation of the field work in Arizona in 1896, the results of 
which have been given in the preceding pages, the author again vis- 
ited this Territory in 1897, remaining about three months, from June 
25 to September 30. 

He was accompanied, as in 1896, by Dr Walter Hough, of the 
National Museum, who rendered most valuable aid, and also by Mr 
F. W. Hodge, of the Bureau of American Ethnology, who joined the 
party at the close of July, remaining with it during the visit to the 
Hopi Snake dances in the following month. 

He was aided also by Mr Frank Zuck, of Holbrook, Arizona, and 
a number of young men from Taylor and Snowflake who were 
employed at Four-mile ruin as laborers. In the Pueblo Viejo Mexi- 
can workmen were relied on, all of whom performed their duties very 

The collection obtained in 1897 was smaller than that made in pre- 
vious years, but it was more varied and more instructive in its bear- 
ings on questions of the migrations of the prehistoric people of Arizona 
than any other thus far made. The ethnological side of the work was 
not neglected. Dr Hough continued his studies of ethnology inaugu- 
rated in the previous years, and has already published the new mate- 
rial obtained by him in the American Anthropologist. a 

As the author visited Tusayan at the time of the Snake dances, he 
made new observations of the Mishongnovi variant of this ceremony. 
A record of his studies of this subject is found in the Nineteenth 
Annual Report of the Bureau of American Ethnology, with accounts 
of the Flute dance at the same pueblo, compiled from notes made in 

On his return from the Southwest the author prepared a prelim- 
inary account of the archeological work, and it was published with 
illustrations in the Smithsonian Report for 1897. This account was 
limited, and contains only the more salient results of the work. The 
present report is intended to be more complete, but is by no means 

The primary aim of the expedition was the collecting of specimens. 
To accomplish this the author was obliged, because of limited forces, 
to spend much of his energy, to the neglect of other lines of investi- 

aV.8, May. 1897. 

fewkes] EXPEDITION OF 1897 121 

gation. The collections of 1897 number a few short of 1,000 entries 
in the catalog of the National Museum. In gathering this material 
the greatest care was taken to label it properly. Neglect of this 
obvious duty has destroyed much of the intrinsic value of many col- 
lections, and has led to errors in conclusions which might readily have 
been avoided. 

The present report completes the record of notes and other data 
bearing on the collections made in the three years during which the 
author has had the honor to direct field work in the Southwest for the 
Smithsonian Institution. There are many obscure points touched 
upon which would be greatly illuminated were it possible to continue 
this line of investigation. So closely connected, however, are the 
archeological and ethnological problems of the Southwest that the 
former can not be exhaustively treated while the latter are so imper- 
fectly solved. 


The summer's field work of 1896 verified by archeological evi- 
dences the truth of the statements of the Hopis that some of their clans 
once lived at Homolobi on the banks of the Little Colorado, not far 
from Winslow, Arizona. It was desirable to study several other 
ruins on this river or its tributaries, and to compare objects indicative 
of the culture of their ancient people with those of this undoubted 
home of early Hopi clans. The author therefore examined ruins 
near Pinedale, on a small southern tributary of this stream near its 
source in the foothills of the White mountains. While employed at 
this ruin he heard of an extensive, undescribed ruin near the Mormon 
town Snowflake, situated on the same stream as Pinedale, but farther 
north. These ruins at Pinedale and Snowflake are almost on the 
meridian of modern Walpi and the mouth of the San Pedro river in 
the Gila valley. 

There is historical evidence that at one time the Hopis used a 
southern trail from their pueblo to the Gila, penetrating to the ranch- 
erias of the San Pedro, and that this trail was rendered impassable 
by the incursions of hostile Apaches in comparatively late historical 
times. An examination of old pueblos situated on or near this 
trail was believed to have considerable importance in connection 
with legends and with historical evidences that it was used by pueblo 

Having studied the archeology of the ruins on southern tribu- 
taries of the Little Colorado, the author made Ids way south of 
the White mountains to that part of the Gila valley which is locally 
known as Pueblo Yiejo, an archoologically uninvestigated region 
which was formerly densely populated and extensively farmed. lie 
desired to discover t he relal Lonship of i he former people of this valley 
with those of the Little Colorado, as well as with those of the Gila 
and Salado rivers, near Tempe and Phoenix, lie likewise wished l<> 


trace the similarities, if any, of the art remains .of these ancient 
farmers with those of peoples who once lived on the Little Colorado 
and its southern tributaries — what resemblances there were in 
implements, pottery, and other art products, and what likeness in 
manners and customs, as indicated by archeological data. 

One of the most important objects of the expedition was to add to 
the sum of available paleographical material from different sections 
of the valley drained by the Little Colorado river. Although the 
amount of this material now in museums is large, an increase of it 
was considered desirable. It has been pointed out elsewhere that 
pictures on old pottery are objective expressions of religious sym- 
bolism, and that they should be treated as such. Each ruin has 
its characteristic designs, and there are features peculiar to certain 
localities. An interpretation of this highly interesting pictography 
can be facilitated by the discovery of new pictures, and the more 
numerous the localities from which it is obtained the more important 
will be its teaching. A discovery of the geographical limits of the 
same S} r mbolism is important, and its connection with the migration 
of certain clans is significant. 

The influence of environment on ancient pottery is a subject of no 
less interest than that of its symbolism. As we pass, in the South- 
west, from one locality to another, the ingredients of the clay from 
which pottery is made change, and the action of fire upon these com- 
ponent^ leads to modifications in their colors when they are used in 
decoration. It would be instructive to follow these changes in their 
many modifications and determine what relations exist between the 
distribution of various clays and different colored pottery. This 
would require a collection of ceramic wares from many localities. We 
can rely only in part on classifications of pottery based on colors as 
indicative of kinship. People of different stocks make pottery of the 
same colors when they use the same or similar clays. Decorations of 
the same kind, or an identical symbolism, are a much more trustworthy 
basis of classification, although not always reliable. 

The ruins studied in 1897 were chosen with a view of obtaining 
comparative data concerning pottery and its decoration from localities 
in different latitudes of Arizona as nearly as possible on the same 
meridian as those studied in previous years. 

The investigations at Kintiel were espeeialby directed toward the 
future plotting of an archeological meridian through Zuni as a basis 
of comparison with the Tusayan zone, in which Walpi is situated. 

It is possible for the expert student of modern pueblo pottery to 
determine at a glance the pueblo in which any piece was made. 
Thus, no specialist would mistake a Zuni vase for one from Acoma 
or confound a Tusayan food bowl with one from Laguna or Santa 
Clara. This exact knowledge has become possible from the fact that 
our museums arc rich in modern ware and familiarity with its char- 

fewkes] FLAN OF 1897 EXPEDITION 123 

acteristics is possible; but an ability to identify modern pueblo pot- 
tery by its symbolism is of little help in the determination of ancient 
ware from the several localities. To determine whether an ancient 
vessel came from near Zuni or from the neighborhood of Walpi we 
must study typical collections of ancient ware. From investigations 
thus far conducted the author is able to distinguish ancient Hopi 
from ancient Zuni pottery, but the geographical limits of each are 
unknown to him and he is wholly unable to distinguish ancient Jemez 
ware from that of Acoma, Sia, or Cochiti. For a provisional classifi- 
cation the author has divided the pueblo area into a number of par- 
allel zones extending north and south. Tusayan lies in one of these 
zones, Kintiel and Zuni in another. Whatever zones it may be nec- 
essary to make to facilitate the study of ancient pottery of the eastern 
pueblos does not concern the present report, but it is worthy of note 
that thus far ancient material from them is so limited that even a 
provisional determination of these areas is premature. 

The author has collected no legends of the Hopi Indians which 
refer in any way to the ruins excavated in 1897, and he believes it is to 
the Zunis rather than to the Hopis that we should look for traditional 
accounts of them which may still survive. Nor has he found anj^ his- 
torical reference to old houses on the Little Colorado river, although 
Four-mile ruin is situated west' of Zuni, and may have been one of 
the pueblos of the Cipias, a sedentary tribe mentioned b} r Spanish 
writers in the seventeenth century. Though this name is said still 
to survive in Zuni legends, the author has thus far failed to elicit 
any information in Hopi stories regarding the ancient Cipias a (see 
page 23), 

The pueblos in the region south of Holbrook are too far east to be 
referred to the Patki and other clans which claim Homolobi as their 
former home; and their surroundings do not in any way agree with 
the current Hopi account of Palatkwabi, the " Old Red land," or the 
"Giant Cactus country." It is always to the mountains south of 
AVinslow that the old men of the Patki clan point when they tell of 
the place of origin of their forefathers. It is instructive to remember 
that the invasion of the Apaches, directed against the modern Hopi 
pueblos, was always from the south, while that of the Utes was from 
the north. The earliest historical account of the contact of the Apa- 
ches with the Hopis indicates that these Athapascan nomads shut 
the latter off from their southern kindred by occupying the trails 
to the Gila and causing Homolobi to be abandoned, and then pressed 
north against the modern towns. 

«Tli<- Hopi name of the modern pueblo Lsleta is Tcipiya. 


The Ruin and its Cemeteries 

Up to within a few years, especially since the American occupation, 
a ruin called Pueblo Grande has been designated on most maps of 
Arizona and New Mexico. This pueblo lies about 25 miles north of 
the railroad station of Navajo, and ten years ago it was one of the 
best ruins of the Southwest, approaching in the perfection of preser- 
vation the famous ruins of the Chaco canyon. To the Navahos who 
range that region the ruin is known as Kintiel, or Broad house. On 
the author's first visit to it, ten years ago, the Avails stood higher than 
a man's head, and the rooms were probably in about the condition in 
which they were shortly after its abandonment. At present very little 
of the ancient walls remains, for they have been torn down by a 
trader, who has used the stones of which they were made in building 
a house and store in about the center of the ruin. In fact, where the 
foundations of the walls of this fine ruin once stood, nothing now 
remains but a trench, for the lower courses of stones, being the largest, 
were sought out for building material in preference to the smaller 
stones which were placed upon them. 

The documentary history of Kintiel is a short chapter. Early 
Spanish accounts do not mention the place, and the Spanish name 
Pueblo Grande appears only on later maps of the country. There is 
said, however, to be a legend concerning it among the present Zuiiis, 
which is mentioned in the Fourth Annual Report of the director of 
the Bureau of American Ethnology (1883) : 

Pending the arrival of goods at Moki, lie [Mr Cushing] returned across the 
country to Zuni, a measure . . . enabling him to observe more minutely than on 
former qccasions the annual sun ceremonial. En route he discovered two ruins, 
apparently before un visited, both, according to Zuni tradition, belonging to the 
Hle-etakwe, or the northwestern migration of the Bear, Crane, Frog, Deer, 
Yellow- wood, and other gentes of the ancestral pueblo. One of these was the 
outlying structure of K'in'ik'el, called by the Navajo Zinnijinne and by the Zuni 
Heshotapathltaie . 

It is interesting to note that all the above-mentioned Zuiii clans 
have or had representatives in the Hopi pueblos, and that at least 
three of them, viz, Bear, Deer, and Yellow-wood, which is probably 
the Hopi Kokop clan, are reputed by the Walpi traditionalists to 
have come into Tusayan from the East. Whether or not these fam- 
ilies of eastern origin are descendants from Kintiel people is impossible 
to say, on account of the author's unfamiliarity with the migration 
legends of these particular clans. It is instructive to learn that with 
the exception possibly of the Frog clan no Patki or Rain-cloud people 
have yet been mentioned from Kintiel, nor do any of the traditions of 
the Patki people mention Kintiel as their former home. 

fewkes] KINTIEL KUIN 125 

No further mention of these ruins is known to the author until the 
description by Victor Mindeleff, in his very important account of 
Tusayan and Cibolan architecture, published in the Eighth Annual 
Report of the Bureau of American Ethnology. 

Fortunately for science, Mr Cosmos Mindeleff camped at Kintiel 
ruin before its destruction and made excellent photographs and plans 
of the ruin, lie likewise conducted limited excavations, which were 
later recorded in a report of the Bureau of American Ethnology. 

Recognizing, on his arrival at Kintiel, that it would be impossible to 
add much to what had been recorded in regard to a ruin so mutilated 
as Kintiel now is, the author naturally sought to learn what he could 
from excavations. The results were somewhat disappointing, and, 
as compared with the collections made at other ruins, only a few 
specimens were obtained from this large pueblo. 

There is one feature in the architecture of the walls of Kintiel which 
seems worthy of special notice, a feature which Nordenskiold recog- 
nized in Mesa Verde ruins, and which the author has described in 
the round house near Montzeimer's ranch, a viz, the difference in size 
of the building stones in the walls and foundations. The largest stones 
occur at the base, or in the lower courses, the smaller in the more 
elevated portions of the walls. This arrangement has a wide distri- 
bution in other parts of the Southwest. 

The nearest point on the railroad to Kintiel is the station Navajo, 
from which there is a good road to the ruin. This road passes in sight 
of several small mounds with indications of former houses, and not 
far from Navajo station there are several ruins, some of considerable 
size, but all in a poor state of preservation. All of these are here 
referred to Zuni rather than Hopi clans, for the fragments of pottery 
which were collected on them resemble the pottery of ancient Zuni 

The exact lines of demarcation between ancient Zuiii and Hopi ruined 
pueblos will probably be impossible to find, mainly because there is 
little doubt that the distinctive features between Zuni and Walpi, so 
marked in modern times, did not exist in ancient times. Clans from 
certain pueblos now in ruins in this region sought union with the 
population of Zuni ; others went to modern Tusayan and were incorpo- 
rated into the population of the villages there. Other families drifted 
out of Zuni and founded pueblos of their own or halted in their migra- 
tion from Cibola to Tusayan and erected pueblos which were aban- 
doned after a few years or generations. 

Kintiel may be classified as a circular ruin (see plate Lin). This 
form is unlike that of any Tusayan ruin, with possibly the exception 
of the two mounds called Kukutcomo, above Sikyatki. Hound ruins 
are foreign to the Hopi country and are absent, from all the portion 
of Arizona south of the presenl Inhabited pueblos of the Hopi reser- 

a Journal of American Ethnology and Archaeology, \. I. L891, p. L27. 


ration. The same may be said of round rooms or kivas. When, how- 
ever, we enter the Zuni belt, which extends from the San Juan and 
Mancos canyon cliff houses south through Zuni, we find both circular 
and rectangular ruins, with circular rooms especially noticeable in 
the cliff houses. Kintiel is one of these, and architecturally, there- 
fore, belongs to the Zuni series, as its geographical position and pottery 
clearly indicate. 

Kintiel is not, however, perfect! j 7, round, but is broader than long, 
assuming a shape comparable with that of a moth with extended 
wings. The two sides were built on sloping land, and between them 
there runs a depression corresponding to the body of the insect of our 
comparison. This median depression is at right angles to the broad- 
est part of the ruin, and in it is the spring which furnished the water 
supply. The present occupant of the ruin, an Indian trader, has 
erected his buildings within the inclosure of the ruin near this depres- 
sion, and has dug out the ancient spring, which furnishes abundant 
water for his purposes. In excavating this spring he found the 
inclosing walls still intact, with a flight of stone steps by which the 
ancients once descended to the w T ater. Notwithstanding sanitary 
objections to such a position for the spring, especially when the 
population of the surrounding houses was large, from a defensive 
point of view it was perfect. The violation of sanitary laws among 
the modern pueblo peoples implies that among the ancients there was 
little regard paid to health in the choice of a water supply, and little 
care in keeping the water pure. 

Extensive excavations at Kintiel revealed a cemetery on the eastern 
side of the northern section. The burials were made close up to the 
outer walls of the buildings, as at Homolobi, but no uniformity in the 
orientation of the bodies was noted. No undoubted evidences of cre- 
mation were detected, and all skeletons exhumed were from subur- 
ban cemeteries. A limited number of specimens of mortuary pottery 
was obtained in the neighborhood of these skeletons. Many of these 
specimens were broken, but others were whole and in good condition. 

The author is inclined to regard Kintiel as a comparatively modern 
pueblo, one of many which were founded later than the earliest Spanish 
invasions. One reason which led to this conclusion is the fine pres- 
ervation of the buildings. Up to within a decade they had hot the 
appearance of antiquity which old ruins always show, nor are there 
now any large refuse heaps or pottery burning places, which so often 
indicate great age, about it. The few graves in the cemeteries and 
the distance apart of those which do exist may be regarded as negative 
evidence of limited value, for it may be said that we may not have 
happened upon the populous graveyards. Yet much more earth was 
moved in the excavations than at Homolobi, and only a tenth as many 
interments were brought to light, and the natural inference is that 
the pueblo was not old. Nothing, however, indicative of white men's 

fewkes] KINTIEL RUIN 127 

influence was found in the ruin. Although it may have been inhab- 
ited since the discovery of Arizona, there is no evidence that Spaniard 
or American ever visited it while inhabited. 

There is a close resemblance between Kintiel, as it was ten years ago, 
and the Pueblo Bonito and other great houses of the Chaco canyon, 
and from its general appearance as compared with these the author 
believes that it is of about the same age. 

There is a Navaho t radition that at least one of the Chaco ruins was 
built by Zuni clans, which would indicate a reason for the similarity in 
the construction of Kintiel or Pueblo Grande and its namesake in 
the above canyon. a Kintiel also resembles architecturally the well- 
preserved Zufli ruins at Archeotekopa, which is described elsewhere, 6 
but we need much more information about these interesting ruins, 
especially about their pottery and the Zuni legends concerning them, 
before it is possible to form any trustworthy conclusions. 

Kintiel is situated on the Leronx wash, c which flows north of 
Holbrook, and turning south empties into the Little Colorado west 
of the town. The wagon road goes from Holbrook past the X ranch 
up the wash to the ruin of Kintiel. There is a ruin of some size at 
Tanner's spring, from which place the author has seen several fine 
specimens of pottery. As these fragments closely resemble the Kin- 
tiel potte^, it is probable that the ruins belonged to the same or 
to closely related people. 

There is also a ruin of some size near HubbelPs store, at the Pueblo 
Granado, about the same distance north of Kintiel that Navajo sta- 
tion is south of it. Pottery from this ruin is ancient, much older 
than that from Kintiel. West of this ruin, at Eighteen-mile spring, 
there is a circular ruin which must also be referred to the Zuni belt. 
The author has been told that there is a Spanish inscription of the 
seventeenth century not far from this spring, but he has never seen it. 

The number of rooms at Kintiel would lead to the belief that the 
population was large, certainly reaching into the hundreds. There 
were evidently several clans living there, and at the lowest estimate 
we are justified in believing that 300 people found shelter within its 
walls. Probably the population was nearer 500 souls, or about the 
same number that formerly lived at Sikyatki. 

A small stone inclosure, apparently a shrine, was found a few feet 
from the outer wall of Kintiel, on the south side. Its contents were 
carefully gathered together and added to the collection. The objects 
found in this inclosure consisted of a number of curiously formed 
stones and concretions, any one of which might, from its odd shape, be 
regarded as a fetish. Several of these stones were rudely worked into 

<«Tho name Kintiel, or Broad House, is applied by the Navahos to at least two circular pueblo 
ruins in tin- Southwest. One of these is in the Chaco canyon, and is said also i>> have been con- 
structed by the Zufiis. 

''Journal of American Ethnology and Archaeology, v. 1. 1801, p. 123. 

Named from the famous guide and trapper Leroux, whose knowledge of the Soul hwesl was 
of such great value to early explorations in this region. 


animal shapes, with head, eyes, and month represented. Similar col- 
lections of stones are common near the approaches to the modern 
Hopi towns and are ordinarily called shrines of the god of death, 
Masauii. It is customary for a Hopi Indian, on approaching the 
pueblo, to throw on these piles any small stone he may have found, and 
in much the same way, no doubt, the pile of stones found at Kintiel was 
formed, for this same custom of casting stones in a pile exists at 
Zuiii, the pueblo to whose people those of Kintiel were allied. 

Just south of the two standing sections of wall there was a cluster 
of stone cysts, probably ancient ovens. They varied in size from 1 
foot or 2 feet square to larger dimensions — 6 by 3 feet. Charred 
wood and ashes were found in some of these, and the bounding stones 
showed the action of fire. These structures reminded one of the 
suburban, communal ovens, which have been described a in the Zuni 
ruin, Heshotauthla. The communal ovens at the latter ruin, like 
those at Kintiel, are situated just outside the walls of the pueblo, 
but unlike them they are, as a rule, round, and of eqilal size. It 
would appear that cooking was done in these ovens rather than in 
the dwelling houses. The Hopi food called pikame, made for cere- 
monial feasts, is still cooked outside the dwellings, and the Zunis 
likewise have ovens separated from their houses, as is common in 
Mexican towns in the Southwest. 

The cemeteries east of Kintiel revealed many skeletons in fair 
preservation, and it was noticed that those near the surface were 
mostly of infants, the adults, as a rule, being found deeper. The 
first skeleton excavated was that of an infant buried under a flat 
stone 2 feet below the surface. The grave had mortuary objects 
in the form of a few miniature rough bowls and a small jar of coiled 
ware. As the excavations penetrated deeper, there were found many 
fragments of pottery, broken ladle handles, ashes, and other indica- 
tions that this was the dump place of the neighboring pueblo, the 
outer wall of which was 50 feet away. 

One of the most instructive burials at Kintiel was found in the east 
cemetery. This was interpreted as a secondary interment. It con- 
sisted of human bones stripped of flesh and deposited in the earth 
with customary mortuary vessels. The reason for the belief that 
these bones were not covered with flesh when the bowls were placed 
upon them is that their position was not that which they would have 
had if articulated. The femurs were placed in the reverse of the 
natural position, and a humerus was found crossing the femur. No 
skull or pelvis was found in the grave. A flat earthen disk was luted 
to the neck of a vase placed on the bones, and there was a food bowl 
near by. 

"Journal of American Ethnology and Archaeology, vol. 1, 1891, p. 133. 

fewkes; KINTIEL RUIN 129 

Pottery from the Ruin 
general features and form 

The pottery from this ruin belongs essentially to the Zuiii type, 
and is very different from that of the Tusayan series. It is, as a rule, 
of coarse texture, and decorated with rude symbols. We miss in it 
the fine yellow ware for which Tusayan is famous, and find in its 
place abundant red pottery, with a comparatively large proportion of 
black and Avhite. The decorative designs are mainly geometrical, 
and picture writing is A r ery limited in quantity. The decoration is 
essentially different from that of Sikyatki, and resembles closely that 
of Ileshotaiithla and Ilalona, two ruins near Zuiii pueblo. 

The greater the number of ancient Zuni pottery objects which 
were examined, the stronger became the belief that the ancient potters 
of this region were inferior to the ancient Hopis in their ceramic 
productions. Modern Zuiii ware is certainly as fine as modern Ilopi, 
and, a priori, the author sees no reason why the older pottery of one 
region should be inferior to that of the other. He formerly supposed 
that this inferiority was due mainly to imperfect collections and that 
the best examples of ancient Zuiii ware were still under ground, so 
that the known specimens gave an imperfect idea of what other and 
larger collections might reveal. 

While these earlier conclusions may be verified by later studies, 
the author now inclines to the belief that the Zunis never advanced to 
the same perfection in the ceramic art as did the Hopis. It must be 
confessed, however, that pottery has been taken from the cliff houses 
north of Zuni which is as fine as the Tusayan ware, and if this excel- 
lent pottery is classified as Zuiii ware, an unfavorable criticism is 
not just. So far as texture is concerned, the Tusayan ware is superior 
to all others in the Southwest, with the exception of the black and 
white ware of cliff dwellers. In the character of designs the superi- 
ority is even greater. In the evolution of Pueblo decoration the 
development of ornamentation advances from geometrical patterns to 
rude picture writing, and, as a rule, the pottery on which the former 
predominate is inferior to that on which the latter is most prominent. 
Not that we should regard this a hard and fast law without exception; 
the cleverest potters often adorn their wares with simplest patterns; 
but in a ruin where most of 1he potter} 7 is decorated with geometrical 
figures, and the few existing pictures of animals — as birds, reptiles, or 
human beings — are rudely made, the artistic development is inferior 
to that where the conditions are reversed. 

Judged by the criterion of designs, Tusayan decoration of ceramic 
ware shows a superiority over all others in the pueblo area, as any- 
one will confess who impartially examines large collections from differ- 
ent areas of the Southwest. 

It would appear, too, that this high development was autochthonous, 

22 eth— 04 1) 



[ETH. ANN. 22 



Dipper with decorated handle., 
from Kintiel. 

and originated within the limited area in the midst of which the 
present Hopi villages are situated, where not only the decoration, 
but also the ware itself is superior. Strangeby enough, the more 
ancient the ruin is, the better is the pottery. This may have a bear- 
ing on certain theories regarding the ancestors of the Hopis, for we 
have been accustomed to hear them spoken of as rude Shoshoneans 

akin to some of the lowest tribes of 
the Rocky mountains, who have 
adopted a pueblo life after they came 
into the pueblo area. The author's 
researches show that only a small 
part of them claim to have sprung 
from the north, but from whatever 
source they came, and whether they 
adopted the pueblo life after their 
arrival or not, they reached a higher 
culture, judged by artistic excellence 
of pottery, than any other pueblo 

Some of the specimens of coiled 
ware from Kintiel are remarkably fine. One of the best is almost 
black, as though discolored by constant use in the fire, and was 
evidently a cooking pot. 

The accompanying figure (75) of a dipper from Kintiel might readily 
be mistaken for a like object from the cliff houses of the Mancos 
canyon. It is a common form of 
black and white ware almost uni- 
versal throughout the Southwest. 
The forms of pottery from Kintiel 
are not exceptional, for all the types 
which were found there occur else- 
where. The rough ware, universal / 
in the pueblo area, is abundant in (* 
the Kintiel graves, and leads all oth- 
ers in number of specimens (see fig- \, 
ure 76) . This is in marked contrast 
to collections from Sikvatki and the 
Little Colorado ruins, where smooth 
decorated ware predominated. 

There were comparatively few fig. 76. 
food vessels, and no large vases 
were obtained. Cups, ladles, vases, and slipper jars were the most 
common pottery forms. A three-lobed cup of red ware was dug out 
of the eastern cemetery. This form is exceptional in the pueblo 
ruins which the author had previously examined, but beautiful spec- 
imens have been found at Homolobi and Chevlon. 

Coiled vase from Kintiel (number 




Fig. 77. Two-handled bowl from Kintiel (number 


The amphora form of globular vessel is rare in Southwestern ruins, 

but is represented by a single specimen (figure 77) from Kintiel. This 

vessel is of black and white 

ware, and the design on the 

equatorial region is charac- 

A very good specimen of 

globular form (figure 78) 

was found at Kintiel. This 

was made of black and Avhite 

ware, and is one of the finest 

specimens in the collection. 
All these examines are 

white ware decorated with 

figures in black, and the 

white is a slip rubbed over 

coarser clay. In firing, since the contraction and expansion of this 

slip is not the same as that of the base on 
which it is laid, we find a crackled surface 
unknown in true ancient Hopi pottery. 

Many of the ladle handles were perfo- 
rated with rows of holes ; several were dec- 
orated with alternate parallel and longi- 
tudinal bands, a type of ornamentation 
which is found as far south as the northern 
border of Old Mexico and has been re- 
corded from Mexican ruins in Chihuahua. 

Several fragments of the necks of vases with pits or depressions were 

found. Some of these pits resembled small cups, 

but the author believes the depressions are finger 

holds, by which the vessel was carried. Fragments 

with similar depressions are found elsewhere in the 

Southwestern ruins, and there are one or two com- 
plete vases with the same finger holds, in which 

there can be no doubt of their use. 


The limited number of specimens of pottery from 
the ruin makes it necessary to speak of this asped 
of the subject in a very general way. 

There is little similarity of piel lire design between 
these specimens and those of modern Zurii which 
have been examined, save in geometrical patterns; 
so that the author is led to suggest a theory to 
account for this fact, similar to that which he has elsewhere advance* I 
to explain the change in symbolism in Hopi poll cry. The differences 

Fig. 78. 

Globular bowl from 

PIG. ?•). Handle of 
dipper from Kintiel. 



[ETH. ANN. 22 

Fig. 80. 

Frog design on bowl froni 

between modern and ancient Ilopi ceramic designs are due to the 

advent of new clans as colonists, for these new arrivals introduced 

their strange ciiltus, of which, up to that time, the Hopis were igno- 
rant. Possibly a similar explanation 
may account for some of the designs on 
modern Zuni pottery. Modern vessels 
from these two regions bear, however, 
widely different decorations. The many 
likenesses between ancient Zimi ware 
and that of Kintiel are the main reasons 
for his association of the two, but these 
similarities are mainly in geometrical 

One or two specimens of pottery from 
Kintiel had handles decorated with the 
forms of animals, and one of these, of 
black and white ware, was particularly 

well made (see figure 79). The intention was evidently to represent. 

some many-legged animal, combining painting with sculpture. 
A knob on one vase has been ident- 

ified as a representation of the head of 

an antelope. This is an unusual form 

of decoration. 

The modification of the handle of a 

dipper into an animal form is not rare 

in ancient pueblo pottery, and the 

author has seen specimens in which a 

mammal, possibly a bear, was represented in that way. In modern 

pueblo pottery, animal forms are very common, and they are espe- 
cially abundant in modern Zuiii ware, 
as an examination of the rich collection 
in the National Museum will demon- 
strate. This method of ornamentation 
is not very common in potter} 7 from 
ancient or modern Hopi towns, though 
the Hopi priests called "mudheads" or 
"clowns" are often represented on the 
handles of ladles, and in the large col- 
lections from Sikyatki not a single speci- 
men adorned in this manner can be 

One of the vessels from Kintiel was 

decorated on the interior with what seems to be a figure of a lizard or 

tailed batrachian (see figure 80). The design is simple, and is not 

unlike figures which are found as pictographs in the Canyon de 

Chelly and elsewhere in the Southwest. 

Fig. 81. Food bowl from Kintiel. 

Fig. 82 

Bird design on food bowl 
from Kintiel. 




The decoration on the exterior of the food basin shown in figure 
81 is highly characteristic and markedly different from that on 
Sikyatki pottery. In this specimen the design on the exterior con- 
sists of a number of interlocked S-shaped figures, which are like- 
wise found on the pottery of the Little Colorado ruins. The external 
decorations on the food basins from Sikyatki are, as a rule, rectilin- 
ear, and curved figures are rare or unknown. A very much mutilated 
figure of a bird which decorates a bowl is shown in figure 82. 

The accompanying illustration (figure 
83) gives a good idea of a Kintiel mug of 
black and white ware and the calcareous 
incrustation with which the majority of 
these ancient vessels was covered. This 
mug is decorated with geometrical pat- 
terns, the nature of which may be seen in 
the illustration. Like many others from 
Kintiel, it was covered with a calcareous 
deposit, which can readily be removed by 

One of the best specimens of white ware 
from Kintiel is shown in figure 84. The FlG< 83, Cu 1 p f 7™ 1 Kintiel (num ' 

ber 176811 ) . 

striking feature of this dipper is the form 

of the handle, which is made in imitation of the head of some animal. 
There were several specimens of bowls and other vessels with heads 
of animals, a feature also common in Tusayan ceramics. 

Miscellaneous Objects from the Ruin 

The stone objects from Kintiel are in no respect peculiar, and con- 
sist of mauls, hammers, axes, spearheads, and arrow points. 

A small slab of stone had three 
cavities, arranged in a triangular 
form, in one surface. There were 
several clay disks, some with a cen- 
tral hole, others imperforate. Rect- 
angular gorgets of red stone were 
perforated at one side as if for sus- 
pension. There is also a tubular 
pipe of red stone in the collection. 
Symmetrical spherical stone balls, ranging in size from a marble to a 
baseball, were picked up on the surface. 

No prayer sticks were found in the graves, but in one of the food 
basins there was a collection of several hundred short sections of wood 
about the size of a small lead pencil, and beveled at both ends. These 
were about an inch long, reminding one of sticks called the "frog 
spawn," wooden symbolic objects made in the Walpi Flute and Snake 

Fig. 8-4. Dipper from Kintiel. 


Bone objects — awls, needles, bodkins, and the like — were numerous. 
Bone tubes of different sizes were likewise found, and a small bone 
gouge accompanied one of the skeletons. 

Fifteen well-preserved human skulls, excavated from the Kintiel 
cemetery, were brought to Washington. 


If we compare the Zuiiian and Tusayan meridian zones of ruins 
architecturally, we find that the3 T closely resemble each other, or, if 
there is any one feature which distinguishes them as groups, it is the 
predominance in the former of circular ruins. Circular ruins are 
absent in the Tusayan series, while more than a third of the Zuni 
series of ruins are round, oval, circular, or semicircular — rectangular 
and round combined. The cause of this predominance is unknown, 
for the explanations which have been advanced to account for round 
ruins in the Zuni belt would seem to be equally applicable to the 
Tusayan belt, where round ruins are absent. 

Not far from Kintiel there is a small, well-preserved ancient house 
called by the Navahos Kinna Zinde. This ruin is in a good state of 
preservation, the stone walls rising high above the foundations. 

As seen from one side Kinna Zinde looks like a round tower, such 
as are found elsewhere in the Zuni belt of ruins. A closer examina- 
tion, however, reveals the fact that only one end of this ruin is round, 
the remainder being rectangular. 

The ruin is situated on a slight elevation overlooking a fertile plain. 
Flooring indicative of two stories is visible, and the poles of an old 
ladder by which there was formerly communication from one story 
to another are still in place. These poles were notched for the inser- 
tion of rungs. 

The author was struck with the scarcity of potterj T fragments and 
other refuse in the neighborhood, and it was concluded that this 
building had not been inhabited for any considerable time. It showed 
no signs of age, and probably was contemporary with Kintiel, which 
is a few miles away. Kinna Zinde was possibly only a summer farm 
home, peopled by farmers from Kintiel, comparable with Pescado or 
Ojo Caliente on the Zuni reservation. In winter the inhabitants 
retired to Kintiel, and in summer they used Kinna Zinde as a pro- 
tected outlook over their farms. Its position was well chosen for this 
purpose, and it was abandoned at about the same time as Kintiel. 


There are few remains of ancient pueblos near Holbrook, Arizona, 
and the Hopi trail from that town to Jeditoh valley is not known to 
pass any considerable ruin. The author has always been astonished 
that the fine spring at Bitarhiitce, the Red cliffs, about 40 miles 
from Holbrook, on the road to the Hopi towns, appears never to have 

-f A- 3 

fewkes] RUINS VISITED IN 1897 135 

furnished water to a neighboring pueblo. <* When the Ilopis went 
back and forth to the Little Colorado, in ancient limes, before IIol- 
brook was built, they took the shorter route to Homolobi. That in 
their communication with Zuni they did not use this trail to the river 
is evident, for the Zuni trail strikes the railroad far to the east. 

Both Ilopis and Zufiis in their intercommunication used the trail 
through Kinticl, because there was nothing to invite them any other 
way. The gateway of modern Tusayan to the southern settlements 
was past Big and Little Burro springs, from Homolobi, and if we 
follow that trail we pass man}' ruins, for the simple reason that it 
is the ancient route of migration. Pueblo Indians, in their journeys, 
go from pueblo to pueblo, stopping for entertainment, so wherever 
we find an ancient trail, there we may expect to find at intervals the 
ruins of old villages. 

Objects from Woodruff Butte 

One of the most conspicuous mountains south of Holbrook, visible 
for some distance along the railroad, is a conical butte called the Pieta 
mountain or Woodruff butte. 
It was not the author's good 
fortune to visit its summit, but 
he obtained by purchase a few 
specimens from that place. 

The most interesting Of these Fig85 - Stone birds from Woodruff butte (number 

were two small stone fetishes or 

bird effigies (see figure 85); there were also some pendants, a few 
beads, and other ornaments. The bird effigies were very skillfully 
carved of stone, and were perforated, evidently for suspension. They 
may have served as fetishes, for they closely resemble similar carved 
objects which are commonly sold at Zuiii. 

Ancient Habitations in the Petrified Forest 

Learning that there w T ere evidences of ruins in the famous petrified 
forest of Arizona, near Holbrook, the author made two visits to it /and 
examined a number of ancient mounds within its area. None of the 
ruins which were seen gave evidence of large size or of a considerable 
population. Many fragments of pottery resembling ancient Zuni ware 
were found, and a few stone implements and metates were picked up 
on the surface of the ground, but the number was too small to 
encourage extended excavations in any of the small ruined house 
clusters which exist in this locality. 

An Indian burial was found on the highest point of one of those 
strange hillocks of the "bad lands" in which the forest is situated, 

"The reader is reminded that this report was written in 1898. Dr Hough's Important 'lis 
coveries in this region were made in L901 (see Report of the U. S. National Museum for 11*01, 



[ETH. ANN. 22 

but the skeleton was too poorly preserved to add to the collections. 
There is a large ruin near Adaniana station, and others in the Petrified 
Forest reservation. 


General Plan 

This ruin is situated 1 miles from Snowflake, and about 2 miles 
from Taylor, Arizona, and is one of the largest in the vicinity. It 
had never been visited prior to the author's work there in 1897, and 
no specimens from this locality are known besides those which he 

The ruin is situated on a bluff overlooking a tributary of the Little 


Fig. 86. View of Four-niile ruin from river bed. 

Colorado called Pinedale creek. One end of the ancient pueblo over- 
looks the stream; the other extends along a low crest at right angles 
to its banks. On the northern and southern sides there are narrow 
plains, that on the south being apparently composed of alluvium 
brought down and deposited by the stream, or washed from higher 
neighboring hills hy torrents of rain, which are often very violent 
in this region. The general form of the ruin is irregularly rectan- 
gular, with no well-defined evidences of a central plaza in the western 
part. The eastern region, however, has a flat top with scattered 
rooms, and was evidently well situated for ceremonial dances or other 

The larger population lived in the western part of Four-mile ruin, 
and probably the eastern region was not permanently inhabited. 

fewkes] FOUR-MILE RUIN 137 

Iii this part of the ancient village there were remnants of circular 
rows of stones, which suggested shrines, and certain piles of refuse 
composed in part of ashes, as though remains of fuel used in firing 
pottery. The eastern quarter of the town does not appear to have 
had an inclosing wall, and no signs of kivas or ceremonial chambers 
were detected. It was the only flat place near the pueblo at all suited 
for sacred dances, and it probably was used for that purpose. 

The accompanying cut, figure 86, shows the appearance of Four- 
mile ruin from the bed of the stream, and gives a fair idea of the 
bluff upon which the mounds are situated. The north cemetery is 
situated at the left of the highest point, and the camp of our party 
is seen at the extreme right. 

A Room in the Ruin 

In order to study the architecture of the rooms of Four-mile ruin, 
earth was removed from one of the best preserved and its dimensions 
were carefully ascertained. This room had in the past been washed 
out by torrents of water, and was on that account easy to clear. It 
was situated on the north side of the highest mound, near the line of 
separation between east and west portions of the ruin. 

The floor was found about 7 feet below the surface. It was paved 
with large flat stones, nicely fitted to each other, and apparently set 
in adobe. On the east side there was a raised banquette extending 
across, and corresponding in a general way with the spectator's sec- 
tion of a Tusayan kiva. It resembled even more closely the raised 
floor which the author has elsewhere described in the cavate rooms 
of Verde valley and the cliff-house rooms of the Red Rock country in 
the same valley. 

About midway in the length of the raised portion, near the remain- 
ing floor of the room, there was a small crypt or inclosure formed of 
flat stones set on edge, and similar in form and position to those 
found in the kivas of the cliff palace of the Mesa Verde. The author 
has seen a like structure in San Juan pueblo on the upper Rio Grande. 
In the floor itself there was a depression lined with stone slabs, which 
may have been a fireplace. The top of the banquette was made of 
smoothly worn flat stones, and its side was plastered. Several very 
finely drilled holes penetrated the flags covering the floor and ban- 
quette, the arrangement of which is shown in plate lviii. These 
were about the same siz<* (that of a broom handle) as the symbolic 
opening called the sipapu in the floor of a Ilopi kiva, rt and when the 
first one was found the author was inclined to interpret it in the same 
way. The subsequent discovery of many others left him in doubt as 
to their proper interpretation. 

a The sipapu of a kiva is symbolic of the opening in the earth through which races, in earliest 
times, are said to have emerged from the under world. 



[ETH. ANN. 22 

The structure of the walls was interesting. They were made of 
adobe, but at regular intervals the much decayed remnants of upright 
posts were found embedded in them. These posts (figure 87) are 
thought to be comparable with similar logs used in the construction 
of the adobe Avails of houses in the Gila valley, as described in a sub- 
sequent account of the architecture of the buildings of the Pueblo 
Viejo (page 177). There were no lateral windows in this room, and 
the entrance was probably from the roof, no remains of which were, 
however, discovered. 

In order to determine the number of superimposed rooms in the 
highest part of Four-mile ruin, the author followed the walls down 

Fig. 87. Upright posts in wall at Four-mile ruin. 

from the surface of the main mound, penetrating through two floors 
before he came to the lowest, which rested on the undisturbed soil. 
It may, therefore, be concluded that the pueblo in this part had an 
altitude of three stories, and it is probable that there was still a 
fourth above, the remains of the walls of which the author was unable 
to trace. There were no walls standing above the ground at any 
point on the mounds, and the general appearance of the ruin is that 
of great age. 

As a rule, the oldest ruined pueblos in the open plain are destitute 
of walls standing above ground ; those with high walls are more mod- 

pbwkes] FOUR-MILE RUIN 139 

ern. This is not a universal law, but it can be relied on with fair cer- 
tainty. Ruins on hilltops have, as a rule, higher Avails above ground 
than those in the plains, even when they are of equal age. Cliff houses, 
on account of their sheltered position, preserve their standing walls 
longer than any other type. No doubt one reason why pueblos of 
the plain, especially such as those in the valley of the Little Colorado, 
so seldom have free walls above ground, is their burial by the dense 
sand storms which sweep over them, especially in the spring months. 
The destructive rains in time wash through their roofs, and water, 
making its way into the joints of the upper layers of the walls, causes 
them to topple over, forming debris at their base. These forces take 
time, but, except in those ruins which have walls wholly of stone, the 
most ancient are reduced to simple mounds penetrated by house walls 
which never rise above the surface of the ground. 

Suburban Ovens 

In searching for the north cemetery the author began a trench 
just outside of the bounding wall, and on excavating a few feet 
below the surface found several cysts like those at Kintiel, which 
were filled with ashes and charred wood. These are interpreted as 
suburban ovens. Similar structures w T ere found at Chevlon in 1896 
and it is not improbable that they will later be found in many other 
ruins of the Little Colorado river. 

Many authors have referred to the absence of fireplaces in ancient 
pueblo rooms, and the existence of chimneys in prehistoric times has 
not yet been proved. The discovery of suburban ovens indicates 
that cooking was done in the open, just as is the case with certain 
kinds of food in modern pueblos. 


It was with considerable difficulty that the author was able to find 
the burial places of this pueblo, and some time was consumed in the 
search. In the ruin at Homolobi and on Chevlon creek the inter- 
ments were discovered just outside the outer walls of the pueblo, and 
it was natural to look in these places for burials at Four-mile ruin. 
Extensive trenches failed, however, to reveal any indication of the 
dead in this part of the mounds. No burials were found close under 
the walls. 

In the course of an examination of the level region some distance 
north of the mounds, near the river bank, the author unexpectedly 
discovered a human bone projecting from the soil. This indication 
was sufficient, and systematic work in the vicinity brought to light 
many skeletons and mortuary objects. 

There can not be a doubt that in the time which has elapsed since 
the burials were made the stream has encroached upon this ceme- 


tery, washing away the superficial soil and leaving a great number of 
small bowlders. Digging among these stones was very difficult, and 
many of the burial objects of pottery were broken in extracting them 
from the earth. This part of the stream bed is not flooded except at 
times of freshets, and it is covered with a scanty vegetation, composed 
mainly of small clumps of sage brushes. This vegetation gave indi- 
cation of the existence of graves, for a skeleton was found under 
almost every bush, often buried less than a foot below the surface. 

A second larger cemetery was found on the opposite side of the 
ruin at about the same distance from the houses as was the first. The 
burials at this place were very deep, but the soil was a sandy allu- 
vium in which the pottery was better preserved. As far down as the 
soil was penetrated skeletons and pottery were found. The greatest 
difficulty in getting them was due to the caving in of the embank- 
ments. Most of the finest specimens were obtained at this point, but 
the supply was by no means exhausted. 

The bodies were buried extended at full length, and with no effort 
at a common orientation. Most of the skeletons were poorly pre- 
served, even the larger bones crumbling as they were removed from 
the graves. A number of perfect skulls, including, those of adults 
and children, male and female, were, however, obtained from both cem- 
eteries. No evidence was noticed of an attempt to cover the bodies 
with logs, as was done at the Chaves pass ruins, or with flat stones, 
as was so common at Homolobi. No fragment of a wrapping of mats 
or basketry was found. 

A cooking pot found in this cemetery contained a lump of clay, 
rib bones of some mammal, a stone polisher, and many cedar twigs. 
Within this bowl were two smaller vessels turned upside down. 

Most of the pottery found in the cemeteries of Four-mile ruin was 
covered with a tenacious, white, calcareous deposit, which was easily 
removed by washing. 



Principal Types 

The pottery of Four-mile ruin is essentially the same as that found 
at Homolobi and Chevlon in 1896, and consists of about the same 
proportion of decorated and of rough, coiled ware, the former 

The rough ware differs but little from that of the pueblos already 
mentioned, but there is a great increase in the number of specimens 
of this ware with a smooth blackened interior. The percentage of 
this kind of pottery increases as we go south from the ruins about 
the inhabited villages of the Hopis, and is greatest in the ruins on the 
Gila-Salado watershed. The blackened interior resembles the black 

*^> **!*;>**%»,, 


ware of Santa Clara pueblo, but no vessel was found at Four-mile 
ruin whose exterior was of this eolor. 

One of the kinds of rough ware which is well represented at Four- 
mile ruin is that decorated on the exterior with geometrical patterns 
(see figure 88). The pigment was applied to the rough outer surface 
of the coils. Commonly, however, the interim- was smooth and black- 
ened, as with certain other rough-ware vessels. The predominating 
color of pottery from this ruin was red, and almost all forms were 
made in this color. It is the characteristic color of pottery in the 
Little Colorado ruins, and is found as far south as Pinedale, reap- 
pearing again in the Gila basin. 

Bowls of red ware with black decorations having a margin of 
white occur in many of the Little 
Colorado ruins. Fine vases of these 
colors, in which white predominates, 
especially around the neck, are char- 
acteristic of ruins in this valley; the 
author has found no record of them in 
the neighborhood of the Ilopi towns, 
or south of the Mogollones. A repre- 
sentative specimen of this type is 
figured in the author's preliminary 

i? -ionr> rni> • „ 4. Fig. 88. Ornamented rough bowl from 

report for 1896. I his ware is not as Fom , mile ruin (mim ^ r 177148) . 

fine as the characteristic cream and 

yellow ware of Sikyatki, but is often made of a finely ground clay 
sufficiently well burned in firing to give fine specimens. 

Gila Type 

The characteristic pottery of the Gila valley is a brownish ware, 
ornamented with red, and is very easy to identify. A specimen of 
this ware has been figured in color in a preliminary report for 1897. 
As far as is known, this kind of ware is generally confined to the Gila- 
Salt river basin. In the excavations of the cemeteries at Four-mile 
ruin two specimens of this peculiar ware were discovered, but the 
author docs not regard the adventitious occurrence of these speci- 
mens, so different from the others in the same ruin, as anything more 
than examples of intrusion, and believes that they were brough there 
from a distance. As a rule, there is considerable similarity in the 
coarse types of pottery from Four-mile ruin and from Pueblo Viejo, 
the upper part of the Gila valley, which the author has not regarded as 
illustrating a theory of transportation of specimens; but the sporadic 
appearance of a prominent type of Gila pottery so different from the 
others appears to him to demand such an explanation. We may sup- 
pose that these specimens went over the watershed of the Gila and 
Little Colorado in the packs of traders, or possibly were carried by 
migratory elans. They were not manufactured by the people in 
whose cemeteries they were found. 



There is nothing peculiar in the forms which the pottery from this 
ruin assumes, though there were a few specimens different from any 
yet obtained from the Southwest. One of the most beautiful of these 
was a globular vessel of red ware, with a graceful neck and symmet- 
rical handle. The ornamentation on this vessel was black and glazed, 
the design representing a highly conventionalized bird. This speci- 
men was perfect, with the exception of a small chip in the lip of the 
orifice. Although a long search for the missing fragment was made 
it could not be found. 

An oval vessel with a hollow handle with external opening recalls 
similar objects called canteens in other reports. They were doubtless 
used for transportation of water, and may be classified as a ceremo- 
nial type of pottery. 

An unusual form, seen in figure 89, is shaped like a saucer, and is 

Fig. #9. Small saucer from Four-mile ruin (number 177131). 

decorated exteriorly with an artistic arrangement of triangles in 
black, bordered with white lines. 

Food bowls predominated in the collection, and the majority of the 
vases were small. 


The contribution of the picture writing on pottery from this ruin is 
highly instructive, and connects the people of Four-mile ruin with 
those of Homolobi and Chevlon. As on the pottery from the latter 
ruins, bird figures are particularly abundant, but there are represen- 
tations of human beings, mammals, reptiles, and insects. 

While, however, there is a general similarity between the ceramic 
pictures of this ruin and those of the pueblos mentioned above, this 
resemblance does not extend into details, and the same may be said 
with regard to other pueblo paleography. The picture writing of 
each pueblo has an individuality which seems to indicate that it was 


independently developed, adapting certain general forms or patterns 
to special ideals. The causes of this divergence in the designs on 
ancient pottery are no more comprehensible than the differences in 
the decoration of modern pottery in two different pueblos. Why, for 
instance, should the symbolism of Walpi differ so markedly from 
that of Zuni, when there are so many points in common between 
the rituals of the two pueblos? The differences in the pueblos are 
mainly due to their clan composition, to the relative prominence of 
different families in them. 


The student of the modern Hopi ritual is familiar with the use of 
helmets in ceremonial dances, and the author has pointed out the 
limitation of those helmets to the rites from the advent of the kat- 
cinas at the winter solstice to the Niman, their departure in July. 
Katcina dancers among the Hopis are masked, and they are the only 
masked dancers in the calendar. Studies of the ancient pictography 
from Sikyatki have not revealed a single figure wearing a mask; but 
the majority of the human figures on modern pottery wear masks or 
ceremonial helmets. The interpretation which is advanced for this 
fact is that the ancients in Tusayan were not familiar with masked 
figures, not having them in their rites, but that in the growth of the 
ritual new clans, in modern times, introduced masked katcinas, and 
consequently modern potters now make figures of them on their pot- 
tery. The logical conclusion would be that, if we find in any ruin a 
picture of a masked personage, the inhabitants of that pueblo must 
have seen a katcina. One of the pictures found on a bowl from 
Four-mile ruin leads to the belief that katcinas were known in that 
pueblo, for it represents a masked dancer (see figure 90). 

The design maybe interpreted as follows: The figure is evidently 
intended to be a drawing of a human being. The head has the form 
of a mask, in which are slits for eyes; the knobs represent feathers. 
The three semicircular figures on the lower end of the bodv rcsem- 
ble rain-cloud symbols, and the double row of rectangles with inclosed 
dots recall the symbol at present used by the Hopis to represent an 
ear of maize. 

There is little doubt that the figure shown in plate xxiv represents 
a human being. All parts except the head are recognizable, and as we 
know from another- specimen that ancient Pueblo artists could repre- 
sent a human head very cleverly, we are called upon to explain why 
they substituted for a head the strange device which is found here. 
The possible explanation is that it represents a mask. The designer 
intended to figure a masked human being or katcina. Now, different 
katcinas are distinguished by symbols drawn on their masks or hel- 
mets, consequently the next step is to compare the helmet of the 
masked figure from the Four-mile ruin with those known in the Hopi 



[ETH. ANN. 22 

The author finds one highly suggestive appendage to the head — 
the radiating crest resembles the feathers in figures of a mythical 
conception called Shalako. We have here a picture with a helmet 
adorned with a crest of feathers, recalling a Shalako, which is a Zuni 
as well as a Hopi conception, derived in Tusayan and Zuni from the 
same source, or from some of. the ruins along the tributaries of the 
Little Colorado. The logical conclusion would be that the people of 
Four-mile ruin likewise recognized this being. 

Apropos of the possibility, revealed by this picture of a masked 

Fig. 90. Human figure on food bowl from Four-mile ruin (number 177061). 

dancer, that masked or katcina dances were once celebrated at 
Four-mile ruin, attention is called to the short distance of this ruin 
from a legendary home of the katcinas near St John, New Mexico. a 
Both Hopi and Zuni legends regarding the ancient home of these 
beings cluster so definitely about a ruin near this town that we may 
suppose that the former inhabitants of that mythical place possessed 
a knowledge of the cult. To the lake near by both Zuiiis and Hopis 

"Kothualewu of the Zuni legends; Winema of the Hopi. It would be a most instructive work 
from a mytho-archeological point of view to investigate the antiquities in the neighborhood of 
St John, especially near the lake so often mentioned in legends. 


make pilgrimages for sacred water; here, likewise, they carry prayer 
plumes. The locality is sacred to the priests of the katcina cult in 
both pueblos. The logical implication is that some of their ancestors 
once lived there. 

The distance of the Four-mile ruin from this place so closely con- 
nected with the katcina cult is not as great by many miles as between 
it and Walpi, not much greater than between it and Zuiii; so that it 
is certainly not improbable that the cult which has made its influence 
felt on these modern pueblos should have been practiced in the pueblo 
now called Four-mile ruin. 

Another picture of a human face, body, and arms is also instructive. 
The head of this figure (see plate XXV a) is unlike any other, but the 
appendages are closely paralleled in figures on certain ancient vessels 
from Oraibi. The mouth is represented by a triangle, as is also the 
case in modern Hopi pictures of the sun god. The arms to the elbows 
are raised to a level with the head, which is circular, with two large 
eyes. The two appendages shaped like quadrants are supposed to 
represent feathers. The bowl on which this picture occurs is broken, 
but it is one of the most beautiful specimens of red ware in the 

A rude figure of a quadruped decorated one of the largest food 
bowls found at Four-mile ruin. Designs of this kind are common in 
pictographs, but are rarely present in pottery decoration. It has been 
suggested that this figure was intended to represent a dance figure, 
and that the caudal appendage shows the fox skin which is at present 
almost universally worn by participants in the sacred dances. It is a 
widespread belief among the pueblo people that in early times, more 
especially when the human race inhabited the under worlds, human 
beings had tails. a Perhaps the ancient potter had this myth in mind 
when some of the human figures represented on old pottery were 


One of the best examples of picture writing from Four-mile ruin 
occurs on a vessel of fine chestnut ware not unlike that of Sikyatki. 
The author formerly regarded this as a picture of a reptile, or possi- 
bly of a horned toad, but there are reasons for identifying it as a 
quadruped, possibly the raccoon. 

The general form of this figure is shown in plate LX b. The head 
has a triangular appendage, the throat is spotted, and the jaws are 
armed with teeth. Two eyes are placed on one side of the head, as is 
often the case in Pueblo drawings of animals. The body is crossed 
by parallel and zigzag lines, and in places is decorated with crosses 
and dots. 

The quadruped figures on the exterior of bowls are mentioned later. 

a Many ancient legends refer to the caudal appendages of men in very ancient times, and it is 
sometimes stated that their tails were cut off byacultus hero. These traditionsaronotconflned 
to the Hopis, but are reported from other pueblos. 

22 eth -04 in 



[ETH. ANN. 22 


Figures of birds predominate in the pictography of all the ancient 
pueblo ruins which have been studied. This is true no less of Four- 
mile ruin than of those lower down on the Little Colorado river. In 
their delineations of bird figures, however, the artists took strange 
liberties with nature, representing birds unknown to students of 
ornithology. One of the most interesting of these from Four-mile 
ruin was a toothed bird drawn on the interior of a food basin. That 

Fig. 91. Bird design on food bowl from Four-mile ruin (number 177203). 

this picture was intended to represent a bird would seem to be shown 
by the representation of wings and tail, though but for the latter 
organ it might be suggested with some justice that a bat was intended. 
In all these representations of mythical animals the imagination had 
full sway. It was not the bird with which the artist was familiar 
through observation, but a monstrous creation of fancy, distorted 
by imaginations — real only in legends — that the potter painted on her 
vessels. Hence, we can not hope to identify them," unless we are 
familial' with the mythology of the painters, much of which has 




perished. The comparatively large number of bird figures on the 
ancient pottery indicates a rich pantheon of bird gods, and it is 
instructive to note, in passing, that personations of birds play impor- 
tant parts in the modern ceremonies which have 
been introduced into Tusayan from the south. 

One of the best figures of a bird found at Four- 
mile ruin is shown in figure 91. The various 
organs can be recognized without a detailed 
description, but the form of the wings is some- 
what different from that thus far shown in picto- 

In the next design (figure 92) we have at oppo- 
site angles of a rectangular figure representa- 
tions of birds, alternating with triangles drawn 

on the remaining angles in a characteristic Zuni and Hopi manner. 
This is one of the few figures in which birds are represented by 

Fig. 92. Bird design on 
food bowl from Four- 
mile ruin. 

PlO. 93. Bird designs on food bowl from Four-mile ruin (number L77170). 

The bird design reproduced in figure !)•*> shows a Long curved snout, 
and parallel lines representing feathers on tail and wings. The two 
legs are thrown out of perspective, but so closely do they resemble 


those of some other bird figures that there can be little doubt of their 
homology. In the same inclosure in which the bird is depicted there 
is also a figure of a dragon fly, and outside the inclosure is a picture 
of another bird. This is one of the most interesting avian pictures 
from Four-mile ruin. The representation of tail feathers by parallel 
lines in this figure is corroborative of the same interpretation of 
parallel lines elsewhere shown on ancient Pueblo pottery. The 
form of the head and the long curved beak is common in several 
other pictures of birds, and an effigy vase with beak of a like 
structure is described from Chevlon ruin in the report of the 
expedition of 1806. 

Fig. 94. Bird design on food bowl from Four-mile ruin (number 177173). 

A very highly conventionalized bird figure is shown in figure 94, 
where the different parts are represented by geometrical lines. 


A large and beautiful food bowl of red ware (plate xxv b) found 
at Four-mile ruin, had an iinusual design representing a moth or but- 
terfly, probably the latter, depicted on one segment of the interior. 
In this design (figure 95) two eyes are represented on one side of the 
head, there is a coiled antenna, and the bod} r and the border of the 
wings are marked with rows of dots. These dots are common fea- 
tures in butterfty figures, as may be seen in modern drawings of this 
insect among the Hopis. 





Representations of the feather, often highly conventionalized, are 
very common in the designs on ancient Hopi pottery, and, as the 
author has shown in a previous article, different kinds of feathers 
have characteristic forms. These designs have been detected thus 
far in the ruins about the inhabited Hopi villages, at Sikyatki, Shu- 
mopovi, and Kisakobi or old Walpi. They have not been found, with 
one exception, in the ruins along the Little Colorado river, though 

Fig. 95. Butterfly design on food bowl from Four-mile ruin (number 177110). 

the author lias been able to examine much larger collections from this 
region than from either Shumopovi or Kisakobi. 

One of the feather symbols was shown to be the triangle, a form of 
which is still preserved in the decoration of modern ceremonial para- 
phernalia. This type of feather design seems to be common in the 
Little Colorado pottery, but is more difficult to recognize and is also 
less common here than it is in the highly instructive symbolism of 



[ETH. ANN, 22 


There were many specimens of pottery from Four-mile ruin deco- 
rated with the various geometrical figures so common on all ancient 
Pueblo ware of northern and central Arizona. The types were 
terrace figures, spirals, frets,, bands, dots, bars, and zigzags. The 
proportion of geometrical figures, as compared with representations 
of animals, was large. As we investigate ruins more and more dis- 
tant from those about the Hopi villages, this proportion increases; 
and if we considered geometrical motives as older and simpler than 

Fig. 96. Sun emblem on food "bowl from Four-mile ruin (number 177058). 

figures of animals, it would seem that pottery ornamentation reached 
a higher development in Tusayan, where drawings of animals and 
human forms predominate. 

The geometrical, figures on the outside of food bowls from the ruins 
on the southern affluents of the Little Colorado are more elaborate 
than those on the northern (see plate lxiii). Modifications of the 
broken line, either in spirals, frets, or bands, are common features of 
the ruins in both regions. 

An instructive piece of pottery from Four-mile ruin was a small 


food bowl ornamented on the interior with a ring (see figure 9G), from 
which radiated serrated bars, the significance of which is unknown. 




In the rich collection of Sikyatki pottery the author found a larger 
food bowl, the interior of which was also decorated with a ring, and 
to this ring undoubted feather symbols were added. It may be sug- 

Fig. 97. Bowl with double spiral design, from Four-niile ruin (number 177102). 

gested, therefore, that the serrated appendages to the ring in the 
above-mentioned specimen may likewise be conventionalized feathers, 

and there are other grounds for in- 
terpreting them in this way. 

The small food bowl shown in 
figure 97 is ornamented with an 
exceptional design, a spiral enlarg- 
ing from the center to the rim of 
the bowl. This bowl is interesting 
as the only one of a pronounced 
heart shape. This form of spiral 
is inst i'iictive, showing 1 lie break in 
the line so characteristic of ancient 
Pueblo designs. 

The decoration shown in the ac- 
companying cut (figure 98) is pecul- 
iar, but effective. The two vertical lines on the neck are repeated on 
the opposite side. Similar markings are found on vases, food bowls, 
dippers, and ladles in all Tusayan ruins, and represent feathers. 

Fig. 98. Decorated vase from Four-mile 
ruin (number 177234). 



[ETH. ANN. 22 

The design shown in the next cut, figure 99, is unique among all 
forms of ornamentation known, and its meaning is incomprehensible 
to the author. 

One of the most characteristic designs, with a spiral motive, is 
shown in plate lxiv, which is typical of many figures on Four-mile 
ruin pottery. This design is- characteristic of the Little Colorado 
river ruins, especially on the red ware so common in them. 

The general character of the geometrical ornamentation of food 
bowls may be seen in plates xl-xlii, lxiii. 

Fig. 99. Unknown design on food bowl from Four-mile ruin (number 177126). 


In his account of the ceramic objects found at Sikyatki the author 
has figured some of the more prominent designs from the exterior of 
food bowls and has attempted a discussion of their significance. In 
the abundant material collected from that ruin no specimen was found 
with figures of animals, with the exception of a highty conventional- 
ized bird. Spiral designs were very rare, the main forms being rec- 
tangular geometrical designs with added feathers. In two instances 




there were human hands or animal paws. A dot with parallel or 
slightly radiating lines was a common feature, and the ornamentation 
was, as a rule, confined to zones or limited to one point on the rim. 

The external decoration on food bowls from the Four-mile ruin 
differs greatly from that of the Sikyatki collection. Both rectangular 
and spiral designs occur, and several specimens have figures of mam- 
mals and birds. 

As a rule, the external decoration is continuous on the outside of 
the food bowl, and is not, as is generally the case at Sikyatki, con- 
fined to one portion. Some of the typical forms of external decora- 
tion are shown in plate lxiii. 

In the account of the pottery from Sikyatki attention is called to 
the predominance of straight lines and rectangular figures on the 

Fig. 100. Bear design on exterior of food bowl from Four -mile ruin (number 176999). 

exteriors of the food bowls. Curved lines, and especially spirals, 
were practically absent in this decoration. This is also true of the 
collection of food vessels from Shumopovi, where a considerable 
number were obtained in 1896. Another peculiarity of the external 
ornamentation of Sikyatki pottery is a design in which we have a dot 
from which extend short parallel or slightly divergent lines; these 
have been interpreted as repesenting a nakwakwoci or feathered 
prayer string. 

The external designs on food bowls from the Little Colorado ruins 
have a large proportion of spirals, and thus far there have not been 
found the dot and appended parallel lines mentioned above. It 
seems, therefore, not improbable that this particular form of the 
feather is peculiar to ruins in the immediate vicinity of the present 
llopi pueblos. 



[ETH. ANN. 22 

On one of the food bowls from Four-mile ruin there was a represen- 
tation of a large mammal which calls to mind a bear (see figure 100). 
This is the only instance known to the author of a representation of 
this animal on the outside of food vessels. 

Pictures of birds are found on the outside of several bowls. One 
of the most exceptional of these is the " twin-bird" design (figure 101), 

Fig. 101. Twin bird design on exterior of food bowl from Pinedale (number 176888). 

which represents two birds attached by their tails. These peculiar 
forms are likewise found at Pinedale and other ruins high up in the 
White Mountain reservation. 

The figure of the bird shown in figure 102 is found on the exterior 
of a food bowl from Four-mile ruin, and is one of the few bird draw- 

„ ' — -r- . 

Fig. 102. Bird design on exterior of food bowl from Four -mile ruin (number 177378). 

ings from the outside of a bowl. The manner of representing the 
claws is one often adopted in avian figures. Parallel lines, for tail 
feathers, are repeatedly found in Southwestern pictography. 

On one of the food bowls we find the accompanying symbol 
(figure 103), which reminds one of the modern rain cloud, so promi- 
nent in Hopi symbolism. It has, however, resemblances to the paw 




of the Dear or badger, and from the fact that a mammal identified 
as a bear is found on the exterior of the bowl illustrated above 
(figure 100), it is probable that this symbol likewise should be referred 
to that animal. 

The very chaste form of geometrical decoration shown in figure 104 

Fig. 103. Bear's paw design on exterior of food bowl from Four-mile ruin (number 177277). 

was found on the outside of a food bowl from Four-mile ruin. It is 
a composition of triangles, T-shaped figures, and terraced designs, 
arranged on a shaded rectangle. 


Three types of rain-cloud symbols are used in the modern Hopi 
ritual. These are the rectangle, the semicircle, and the triangle. 
The two former are ordinarily triune. We have either three semicir- 


Fig. 104. Geometrical design on exterior of food bowl from Four-mile ruin (number 177000). 

cles or three rectangles combined, the latter appearing as a stepped 
figure. The number of components may be multiplied, in which case 
we find many semicircles approximated, their shapes somewhat modi- 
fied by the juxtaposition, or many rectangles combined, forming 


It is easy to mention instances of rectangular rain-cloud symbols 
represented in the modern ceremonial paraphernalia. We find them 
carved into tablets on the heads of many dolls and idols. The Humis 
katcina dancers wear them on their helmets. They are painted on 
the uprights of altars, woven into baskets, and embroidered on sacred 
dance kilts. 

This terrace form occurs as a rain-cloud symbol on several bowls 
from the ancient ruin of Sikyatki. It is also found on mortuary stone 
slabs at the same ruin. The four sides of the ancient shrine in the 
Awatobi kiva had rectangular rain-cloud symbols of different color, 
showing that this form was recognized in this pueblo. The author 
has elsewhere pointed out its existence in the Homolobi ruin, and now 
the same type is reported from Four-mile ruin. While, as a general 
thing, this form is the predominating type of rain-cloud symbol used 
in the katcina celebrations, it is not confined to them, but is also found 
in the Flute ceremonies and elsewhere. 

The semicircular type of rain-cloud symbol is no less common than 
the rectangular in the modern decorations, and, while most abundant 
in ceremonials which occur between the departure of the katcinas and 
their advent, it is not wholly absent in the masked dances. This 
form has not yet been found on ancient Hopi pottery— which fact 
leads to a belief that it is of late introduction. It is, however, very 
conspicuous in the ceremonials introduced into Tusayan by the Patki 
or Rain-cloud people, and it is a significant fact that the totemic sig- 
natures of members of this family have the same form. A ready 
explanation of the existence of this motive in Walpi would be that 
the southern clans introduced it, and its occurrence in the Snake 
ceremony would be interpreted as an example of intrusion. 

The semicircular type of rain-cloud symbols is not considered a 
development of the rectangular, or vice versa; but it is thought to be 
a new symbol of foreign origin, the rectangular being the older in 
this particular locality. 

The triangular rain-cloud symbol is less common in modern designs, 
and is rare or unknown in ruins near the modern towns. The majority 
of examples of it come from the Little Colorado ruins, but it occurs 
on some of the idols used in Walpi at the present day. 

A food bowl decorated with triangles arranged in such way and with 
such an association that they may be interpreted as rain-cloud sym- 
bols was found at Sikyatki, but this interpretation is doubtful. 
Another bowl from Shumopovi admits of the same interpretation. 

A symbol of the rain cloud among the people of the pueblo — now 
a ruin — at the mouth of Chevlon fork, was a triangle inclosing a rec- 
tangle. These symbols were found on a stone slab excavated from 
that ruin in 1896, and were figured in reports of the work accom- 
plished in that year (see plate xlvi). A beautiful large food vessel 




dug out of the north cemetery at Four-mile ruin, shown in figure 105, 
is decorated with triangles which are also supposed to be rain-cloud 
symbols. Above them is a semicircular band which is identified as a 
representation of the rainbow. 

An example of the triangular form of symbol representing the rain 
cloud is found on one of the effigies of the Flute altar, and is figured 
in an account of the Walpi Flute observance. a Many of the rattles 
used in katcina dances have on each of their flattened sides four tri- 

Fkj. 105. Cloud emblem on food bowl from Four-mile ruin (number 157352). 

angles united at one angle, and with parallel lines representing fall- 
ing rain on the sides opposite their union. These figures have a 
distant resemblance to feather symbols, as may be seen by comparison 
with some of the bird designs from Chevlon ruin. 

It will be seen from the foregoing account that there are three types 
of rain-cloud symbols in use in the modern Hopi ritual, the semicircle, 
rectangle, and triangle. 

In tin- same way it can be shown thai there are at least 1 wo types 

"Journal of American Folk-Lore, V. 8, n. 27, pi. IT, Hit. 1. 


of sun symbols, and there are other instances which might be men- 
tioned of two or more symbols representing the same thing. This 
duplication is explained by the composite nature of the tribe, one 
family adding one type, another a second, and so on. In the amalga- 
mation of the clans each of these symbols becomes no longer limited 
to the family which brought it to the pueblo. While the semicircular 
rain-cloud symbol predominates, the survivals of the triangular and 
rectangular are numerous and suggestive. 

The oldest form in Tusayan, so far as archeology teaches, is the 
rectangular, but the triangular is possibly equally ancient in the 
ruins along the Little Colorado. 

Character and Treatment of Mortuary Pottery 

It must be confessed that the pottery now placed over the dead is 
of poor quality and scanty in quantity, as compared with that used 
by the ancients for that purpose. The fine ware rarely serves this 
purpose, but is retained in the household. It may be interesting to 
note that among the modern Hopis special pottery objects are not 
manufactured for mortuary purposes, and the same is true of ancient 
burials. In the latter many of these objects show manifest signs of 
former use in the household. 

To what extent the survivors of the deceased purposely broke 
mortuary vessels, nicked fragments from them when they were depos- 
ited in the graves, or in other ways symbolically "killed" them, it is 
very difficult to say. Many mortuary vessels have been found which 
were as perfect as when made; others were undoubtedly purposely 
broken before they were deposited with the dead. The great pres- 
sure of the earth above them doubtless fractured the largest number, 
and manjr were broken while being extracted from the soil. There is 
no direct evidence that mortuary pottery was ever to any great extent 
purposely broken before it was deposited in the cemeteries of Four- 
mile ruin. 


One of the bodies exhumed from the northern cemetery was accom- 
panied by a potter's "outfit," consisting of the different ingredients 
used in making pottery and of smooth stones and other implements 
with which it was made. In other ruins the author has found masses 
of potter's clay such as are used by the potter, but nowhere as complete 
a collection of clays, pigments, and the like as in this grave. 

The objects were: 

1. Knife made of a rib, for cutting clay 

2. Knife made of a rib, for cutting clay 

3. Stone for rubbing, stained green 

4. White clay 

5. Yellow clay 

6. Greenish clay 

7. Micaceous hematite 



Comparatively few stone implements were collected at Four-mile 
ruin, and they were, for the most part, so similar to those from other 
Little Colorado ruins that much space need not be here devoted to 

There were found several serrated stone implements which seem 
worthy of special mention. They are made of hard stone, chipped to a 
sharp, toothed edge. The use of an implement of this kind is obvious; 
for with it, as with a file, a number of mechanical operations, such as 
sawing, filing, and scraping, are possible. Specimens of this form 
occur at almost every large ruin at which the author has worked in 
the last two years, and many of them were picked up from the surface 
of the ground. 

The number of small stones showing pecking or artificial working 
which can be found at a Southwestern ruin is much larger than the 
proportion in collections would seem to indicate. From their great 
weight, as well as their numbers, the majority have to be left behind, 
and as a rule those which are destitute of a special form are rejected. 

It was apparently the Indian custom to pick up any stone near at 
hand, to use it for pounding or other purposes as long as needed, and 
then to cast it away. It thus happens that innumerable stones 
slightly pecked on one or all sides, but without the form of any imple- 
ment, are very numerous upon the mounds of almost every ruin. 

The burials in the north cemetery were deep, and there was evi- 
dence that a considerable quantity of soil had been deposited over 
them, having been washed down from neighboring mounds. A few 
feet below the present surface of the ground in this superimposed 
soil the stone object shown in figure 106 was found, the probable use 
of which was a subject of some speculation. Having occasion later to 
open a room in the mounds above the point where this stone was 
discovered, the author found on the floor, several feet below the 
surface of the soil which filled the room, other specimens having the 
same general shape and character. In a gully between the room and 
the cemetery there was still another of these objects — making in all 
seven specimens. 

The localities in which these stones were found indicated that they 
all belonged together, and that the two found outside the room had 
been separated from the others and had been rolled down the sides of 
the mounds, perhaps by the water, the course of which is marked by 
• leep gullies in their sides. The forms of all these stones are much the 
same, irregular, ovate, with one flat side, and truncated at one pole. 
They were evidently fashioned with care, and, as the rock is hard, they 
must have been made with considerable difficulty. All had a small pit 
or depression on the flat side near the rounded pole. 

Several suggestions were made by members of the party regarding 
the possible use of these stones, of which the following seemed to be 



[ETH. ANN. 22 

the best: That these stones are simply supports for rods used in 
weaving girdles, especially the great white sashes worn by katcinas. 
Identical supports for rods of this kind are used at the present day 
for the same purpose. 

The frequency with which stones used for grinding corn are found 
in graves may be explained by their prominence in the life of the 
women, in whose graves they generally occur. These utensils are 
ordinarily found in a reversed position near the middle of the body. 
The custom of buiying metates in graves is known from the ceine- 



•.*, • •• '..,<:-■ 

Fig. 106. Stone used in belt frame, from Four-mile ruin. 

teries at Sikyatki, Homolobi, Chevlon, Chaves pass, and Four-mile. 
A doubtful instance occurred in the Kintiel burial ground, where a 
metate was found in the graveyard, but not near any skeleton. 



The presence of stone slabs, some of which are of considerable size, 
has been recorded in several ruins of New Mexico and Arizona, and 
these objects were also found at Sikyatki, Homolobi, and Kintiel. 
Several of these specimens were collected at Four-mile ruin. Many 
of the perforated stones were extracted from the floors of the kivas, 




others, generally with an orifice of larger size, from the soil covering 
the rooms. 

It has been suggested that some of these perforated stones were 
formerly built into walls of rooms to partially close the passageway, 


i <j~ 




FiG. 107. Stout: slab from Four-mile ruin. 

but their presence in graves is not readily explained by this theory. 
Their fashioning demanded considerable Labor, and the author recalls 
one of t hese perforated stones where 1 he edge had been worked smool li 
with greal care. 

22 ETii— <)4 



Iii his report on the operations in 189G at Homolobi and the Chev- 
lon ruin, the author called attention to the presence in graves of 
stone slabs on which figures of rain clouds were depicted, and in the 
excavations at Sikyatki he found similarly decorated stone objects. 
The practice of burying stone slabs ornamented with rain-cloud sym- 
bols was not unknown at Foiir-mile ruin, as one of the objects from 
graves at that place attests. This specimen has a rectangular form 
and is decorated with a terraced rain cloud painted in black out- 
line on one side. It is possible that the grave from which this slab 
was taken was that of a priest, and that this object was formerly used 
in ceremonies, as is the case with certain altar paraphernalia of the 
same character in the modern ritual of the Hopi Indians. The repre- 
sentation of the rain cloud on a mortuary stone slab is the expression 
of the idea that the dead become rain makers or rain gods. This form 
of ancestor worship is a highly modified one, which can be directly 

traced to the arid environment in which the 
ancient people lived, and their status as agricul- 
turists, which made rain a prime necessity to 

This slab was likewise decorated with a row of 
triangular markings, and had perforations at the 
corners. A second slab, of less regular form, 
was likewise found at Four-mile ruin, but upon 
it the terraced rain-cloud figures were not as 
Fig. 108. Copper bell from distinctly drawn as on the preceding. There 

Four -mile ruin (number .. _ ., , . ... , 

17780 4) was also found a stone slab with rectangular 

figure of unknown meaning drawn upon it with 
black pigment. A stone slab somewhat like this was found at Sikyatki 
in 1895. 

While strolling over the mounds the author found a slab of stone of 
unknown use (figure 107). It was set upright and photographed. 
The object was about 4 feet long and about 8 inches wide, tapering 
slightly, and smooth on all sides. This slab had without doubt 
been worked into a regular form, and was a lintel of a doorway or 
some other part of a house. 


The occurrence of bells made of copper has been recorded from 
several ruins in Arizona. The specimen obtained at Four-mile ruin 
(figure 108) is in no respect different from those previously mentioned, 
and belongs to the type constantly found in the Gila valley and in old 
Mexico. From the limited number of these bells in Arizonian ruins 
very meager conclusions can be drawn, but the author supposes that 
they were introduced from the south, rather than that they were manu- 
factured by the former inhabitants of the ruined pireblo. There are 
indications of great antiquity in some ruins where they have been 


The ball was taken from the hand of a skeleton exhumed from the 
cemetery north of the pueblo. It was much corroded, and broken on 
one side, and the small stone which served as a clapper had become 
firmly fixed to the inner wall by the corrosion of the copper. 


No fragments of mortuary prayer sticks were found in the cemeteries 
at Four-mile ruin, but this negative evidence docs not prove that they 
w T ere not in use among the inhabitants. The soil is so moist that there 
is doubt if these wooden objects would last long in it, though their 
preservation in the Chevlon ruin, where somewhat similar conditions 
prevail, shows that their absence at Four-mile ruin may furnish posi- 
tive proof that they were not used in burial. 


One of the instructive objects taken from the north cemetery at 
Four-mile ruin was a rattle made of a small gourd. This rattle had 
an oval shape, and was decorated with red and green paint, on 
which was the impression of feathers. The handle, which was broken 
from the rattle, was not found. The occurrence of this gourd rattle, 
identical with those still used in Pueblo ceremonials, gives archeo- 
logical evidence of its use in ancient times, probably as an accom- 
paniment to songs in religious rites. 


Although fully as many skeletons were exhumed from Four-mile 
ruin as from some others, the small number of marine shells, as 
compared with those found at Homolobi and Chaves pass, was notice- 
able. Though the ruin is situated in a latitude south of Chaves, 
only a few fragments of shell were found there, Avhile there were 
several hundred specimens from the latter ruin. This can be explained 
only by the theory that the Chaves pass and Homolobi people had 
more marine shells than those of Four-mile ruin, that they were in 
more direct contact with the ocean, or with people who obtained them 
from the sea by barter or otherwise, thus indicating a direct relation- 
ship between them and peoples of the south. The ancient trade in sea 
shells was along the Gila river, up its northern tributaries, and across 
the Mogollones to the Little Colorado river. Chaves pass was in 
the direct line of this trade; Four-mile ruin was not, and the scarcity 
of sea shells in the latter locality is explained by its distance from the 
sea and the difficulty in reaching tribes nearer the Gulf of California. 

The scarcity of beads and turquoise ornaments in the collections of 
1807 was in marked contrast with the wealth of these objects at 
Homolobi and Chevlon. While this parity may be in part due to the 
limited amount, of soil removed in the work, it must also be remem- 
bered that the pueblos which were excavated in L897 were smaller. 



The bone implements found at Four-mile ruin were similar to those 
collected, in 1896 at Chaves pass. They consisted of awls, needles, 
and bodkins, many of which were made from the wing and leg bones 
of the wild turkey or the tibia? of antelopes. There were also larger 
implements made of the bones of antelope and deer. 


A large collection of animal bones was obtained from the rooms at 
Four-mile ruin, but they have not yet been identified. 


Within a radius of a few miles of Snowflake there are several 
ruins, some of which are of considerable size. The ruin near Shum- 
way is one of the largest of these, and would well repay extensive 
excavations. There are ruins on the opposite side of the creek from 
Four-mile ruin, but these are smaller, and the elevations on which 
they stand have been diminished by deposition of the soil by the 
stream about their bases. The cemeteries have been so deeply buried 
under the accumulated earth that extensive excavation would be 
necessary to lay bare the objects which they contain, and, as the 
mounds themselves are small, the author did not attempt this work. 
The collections made at Four-mile ruin will undoubtedly serve as 
typical of those which could be taken from adjacent mounds, as the 
people of this whole neighborhood were probably in about the same 
stage of culture. 


( The Buildings 

The road leading south from Holbrook to Fort Apache, in the 
White mountains, divides just beyond Taylor, and one division con- 
tinues to a small settlement among the pines, which is called Pine- 
dale. This is a beautiful place to camp, surrounded by high trees, 
is well watered, and in places has fertile stretches of land suitable 
for farms. Two extensive ruins reported to me from that locality by 
Mr Frank Zuck, of Holbrook, promised interesting results if proper 
excavations were made in or near them. 

Accordingly, work was begun, with 5 Mexican laborers, near the 
middle of July, and extensive excavations were made in the larger 
ruin. The results were not as satisfactory as had been hoped, but 
several important facts were brought out by the attempt. A small 
collection rewarded the work at this place. 

The two Pinedale ruins lie on either side of the, road just beyond 
the church of the town, and a few hundred feet from the new stone 
schoolhouse, one of the best in this part of Arizona.] PINED ALE KUINS 165 

Of these two ruins, that on the left of the road is the remains of a 
pueblo of compact form, with a central plaza obscurely indicated. 
From the general appearance of the ruins it is judged that the pueblo 
was at least several stories high, but no sign of wall was seen above 

The ruin to the right of the road covered more ground than the 
other. It was of rectangular form and apparently single storied. 
This ruin was evidently an ancient one, and many tall, fine trees were 
found growing from the soil in the rooms. The walls, however, had 
so fallen in that there were not more than traces of houses to be seen 
marking the former extent of the ruin. There was no evidence at 
any point that the rooms ever had more than a single story; and 
evidences of the gateways entering the ancient plaza were sought in 

The relationship of these two Pinedale ruins to each other appears 
to be as follows: The compact ruin on the left side of the road appar- 
ently contained the greater part of the population, while the rectan- 
gular building served as a place of refuge, for which its mode of 
construction made it admirably suited. If the theory is a correct 
one, it is probable that the rectangular portion was of later date than 
the compact one, and this is also indicated by its general appearance. 

At various localities in the Southwest are found in close proximity 
ruins of buildings which apparently have a somewhat similar rela- 
tionship. Thus in the Tsegi canyon one sometimes finds extensive 
ruins at the base of a cliff, and in the caverns above inaccessible cliff 
houses. Another very good illustration can be seen near Ramah, not 
far from Zuiii, where there is a fine rectangular ruin on the hilltop 
and the remains of an extensive pueblo at the base of the same eleva- 
tion. The more inaccessible of these buildings was probably a place 
of refuge for the inhabitants of the more exposed pueblos in the plain 
and their contiguity made access from one to another easy. The same 
explanation may also be suggested for fortified hilltops near ruins, 
so well illustrated in so-called trincheras of the Verde valley. 

In localities like that of Pinedale, where there were no adjacent 
caverns or hilltops convenient for fortification, a special building 
was erected for defense and refuge. This method was adopted in the 
Buena Vista ruin, situated in the Pueblo Viejo, to which reference 
will be made later. 

It would appear that a specially erected building for refuge and 
another for habitation is a far less practical arrangement for defense 
than a combination of both in one. This has led to the building of 
habitations on inaccessible heights, in caves, or on mesa tops, or to 
the construction of the pueblo in such a form as to make it easy of 
defense Thus, the houses are so placed that the highest wall is on 
the outside, where it sometimes rises to the altitude of several stories, 
sloping toward the middle of the town. Entrance into such a walled 



[ETH. ANN. 22 

town might be either by ladders, which could be drawn on 
the roofs, or through breaks in the walls or gateways. 

The circular form of building would be a natural evolu- 
tion of this form of a fortified pueblo, a survival of a plan 
of encampment adopted by nomadic Indians, as others 
have pointed out? No doubt sociological and other rea- 
sons also played a part in the circular arrangement of 
houses inhabited by different clans, but the princi- 
pal cause was the need of defense. 

. , 

Fig. 109. Bone implement from 
Pinerlale ruin (number 176964). 


From one of the cemeteries of the larger Pine- 
dale ruin several skeletons were exhumed. It was 
situated close to the outer wall, as at Homolobi 
and Chevlon, and the skeletons were accompanied 
with mortuary pottery. There was no good evi- 
dence that the ancients in this pueblo burned their 
dead, and logs or stones were not found over them 
as at Chaves pass. This was surprising in so well 
wooded a region. The bodies were not', apparently, 
wrapped in matting. 

The pottery is essentially the same as that taken 
from ruins farther down on the Little Colorado, 
and the decorative symbols are much the same 
as at Four-mile ruin. With the exception of a pic- 
ture of a bird on the interior of a food bowl, and 
several more conventionalized bird designs on the 
exterior of another, no animal pictures were found. 
The majority of the decorations were of the geo- 
metrical type. Rough ware and decorated pottery 
occurred in about equal proportions. One of the 
most beautiful specimens of red ware obtained 
in 1807 was exhumed at Pinedale. The 
decorations, both interior and exterior, were 
well made and the ware itself was of the 
finest type. Mr Zuck discovered this ceme- 
tery of the ancient Pinedale ruin in the year 
1896, and removed from it several vessels 
which will compare well with any yet found 
in the Little Colorado basin. There is evi- 
dence, therefore, that as we leave the river 
the pottery does not deteriorate. 

Several forms of bone implements were 
obtained from the excavations at Pinedale 
ruin. These came chiefly from the former 
dwelling rooms, and were rarely found in the 
cemcl cries. They consist of awls, bodkins, 


pins, needles, and pointed implements used in weaving and sewing. 
The largest specimens were made from the bones of deer and ante- 
lope ; the smaller, for the most part, from bones of rabbits and birds. 

A bone implement was found in Pinedale rain cemetery for the 
use of which there is no satisfactoiy explanation. It is made from a 
human arm bone, cut off a short distance from the trochanter. There 
can be no doubt that the cut is artificial, as the marks of a primitive 
instrument are visible, while there was evidently an effort to polish or 
otherwise work the surface of a similar specimen. The majority of 
these bone objects were made of the humerus of the wild turkey, one 
only being made of a human bone. 

One of the most exceptional bone objects found at Pinedale was an 
implement with two long prongs, unfortunately broken at the end 
(see figure 109). 

A shallow house-burial in one of the rooms of the ruin to the right 
of the Pinedale road contained calcined bones, evidently human, a 
copper buckle, and a few army buttons showing the action of fire. This 
was evidently an intrusive burial, much later than the others, and 
there is reason to believe that it was made long after the room in 
which it was found had been deserted, though there was no way of 
telling whether the fragments of the skeleton were those of an Indian 
or white man. 

The tall trees growing from the debris filling the rooms of the rec- 
tangular ruin at Pinedale show that the pueblo was of great age. 
Fortunate^, one of these had been sawn down, revealing the number 
of rings indicative of its age. Though it was not possible to count 
these with certainty, over 100 concentric layers could be made out 
without difficulty . In the room where these metallic objects were 
found grew one of the largest of the trees. 

A considerable collection of crania was made at Pinedale, as at Kin- 
tiel and Four-mile ruins. 


During the troubles in the Tonto basin a few years ago, a party from 
the basin visited a ranch owned by a man named Stott, a few miles 
west of Pinedale, and hanged him for alleged horse stealing. The 
ruin called by his name is a few rods from his cabin, now deserted. 
It is a fine ruin situated in a beautiful park of lofty pine trees, and 
offers opportunities for archeological study, but is inconvenient for 
extensive work on account of its distance from a base of supplies, the 
nearest place, Pinedale, affording only a limited supply of provisions. 
In general character this ruin resembles those near Pinedale, and the 
few fragments of pottery which were picked up on t lie surface are 
identical with those from Four-mile ruin, near Snowflake. A mile or 

"The pueblos near haunts of deer and antelope have a larger proportion of bones of these 
animals; those in the plain have more bones of rabbits and birds. The fauna of the region is 
accurately reflected in tin- bones found in its ruins. 


more southwest of this ruin there is still another, much smaller, 
crowning a hill top, with evidences of a considerable former popula- 
tion. Many fragments of pottery were strewn over the surface of the 
ground, and a few foundation walls were traced, especially on the 
highest point of the hill, but none of these rose above the surface of 
the mounds. The general character of all these ruins is the same as 
that of the Little Colorado series. 


The Valley and its History— General Features of the Ruins 

It will be seen b}' an examination of a map of Arizona that the 
ruins at Pinedale and Stott's ranch are very near the sources of some 
of the southern tributaries of the Little Colorado. They are situated 
high up on the northern foothills of the mountain area, the White 
mountains, which high lands constitute the watershed between the 
Gila and Little Colorado drainage areas. Although the distance is, 
comparatively speaking, short in a direct line from the sources of the 
tributaries of these two rivers, the intervening country is very broken 
and in places is impassable. It is especially desirable from an ethno- 
logical point of view to examine whatever ruins may exist in that 
region, since they may be regarded as frontier settlements of ancient 
peoples which, with many points in common, have many differences; 
but the author did not find it possible to do this. 

It was, however, possible to take up the problem whether there is 
a close likeness between the ancient culture of the Upper Gila and 
that of the people who lived near Phoenix and Tempe. The author 
went around the White mountains, via the Southern Pacific railroad, 
and approached the Gila from the south. The section of this valley 
chosen for archeological study is almost directly south of Pinedale, 
and is locally known as Pueblo Vie jo. 

The name Pueblo Vie jo is given to a portion of the valley of the 
Gila from Pima to San Jose, between Mount Graham and the Bonita 
mountains, forming the greater part of Graham county, Arizona. 

This valley was traversed by the "Army of the West" in 1847, and 
the attention of Americans was first called to it by the reports of 
Emory and Johnston, in their " Notes on a Military Reconnoissance," 
published by Congress shortly afterward. These reports mention the 
antiquities of the valley, and have remained for fifty j^ears the only 
available accounts of them. These authors refer to and figure some of 
the characteristic fragments of potte^, and speak of circular ruins. 
No remains of circular buildings can now be detected, and the author 
has grave doubts that the circular form of buildings ever existed in 
this region. The circular structures were more likely reservoirs. 

This valley was probably known to Spanish explorers as far back 
as the seventeenth, and possibly the sixteenth, century. The com- 


monly accepted route of Coronado would have led him to cross the 
Gila not far from the mouth of its tributary, the San Pedro, where 
there was a trail to Moqui, and probably also to Zuni. If, however, 
as is urged by Dellenbaugh, he took a still more easterly route, and 
Cibola was situated near the Florida mountains and not at Zuiii, 
Pueblo Vie jo and the Gila river are far to the west of his route. 

Documentary history of the Pueblo Viejo in the seventeenth cen- 
tury is practically wanting. None of the great Spanish explorers 
passed through the valley in this epoch, when the region was entered 
along the Rio Grande by way of El Paso del Norte. 

In the first decade of the eighteenth century there were apparently 
no rancherias in the Pueblo Viejo valley. The accounts of the several 
expeditions of Garces, and contemporary maps, give no indication of 
inhabited rancherias east of the mouth of the San Pedro, and no 
mention is made in the diary of this devoted priest of people other 
than Apaches living on the upper Gila. But the existence of ruins 
near the mouth of the San Pedro is noted, though it is highly probable 
that they became such long before that time. 

With the advent of Apaches the population of Pueblo Viejo 
retreated to the west, abandoning their farms one after another, 
until they came to the Aravapa canj 7 on. Here they may have inter- 
married with other stocks, and the Sobaipuris of the earl} 7 years of the 
eighteenth century probably contained some of their descendants. 
They or other survivors never returned to their old homes in the rich 
plains they had abandoned. 

Pueblo Viejo was apparently uninhabited by Mexicans or sedentary 
Indians at the time of the passage of the Army of the West, and 
the mounds indicating former houses were frequently noticed at that 
time. Their age was even then a subject of comment. 

The appearance of Pueblo Viejo at this time was probably not unlike 
that of those sections which are not now farmed. A dense growth of 
mesquite and cactus covered a sandy soil, which in the dry season 
turned to dust, covering the traveler or hovering in clouds behind 
him. Most of the larger specimens of mesquite and other trees 
have long ago been cut down, but the great growth which this tree 
may have reached can be judged from a few survivors. In places 
along the bank of the Gila there were clumps of cottonwood trees, 
some of which even now present a delightful sight to the weary 
traveler. In the rainy season the river overflowed its banks, flood- 
ing the neighboring valley for miles. The river, although fordable in 
the dry season, was so swollen after rains in the mountains as to be 

The scenic beauties of the valley have not changed since the 
Indians lived on the Gila banks. The lofty Graham mountain, the 
black sides of which glisten with streams of water, is a beautiful 
sight from almost any part of the middle region of the valley. It 


is covered in places with tall pines and other trees, and is a grateful 
place of resort in the hot summer days. The still more picturesque 
JBonita mountains, with their serrated summits, hem the valle} 7 on 
the opposite side, and north of these is a broken country, almost 
impassable, yet with ruined cliff houses and other evidences of a 
former occupation. 

The many ruins in the Pueblo Viejo are all of the same type, viz, 
clusters of rancherias with a central building which may have served 
as a citadel for defense. Whether any special building was set aside 
for a ceremonial room or temple is an unanswered question, but 
there is some evidence that the central building may have sometimes 
served for that purpose. 

Although a number of clusters of mounds were found in Pueblo 
Viejo, there were two which were specially examined — that at Solo- 
monville, called Epley's ruin, and that at Buena Vista, a short dis- 
tance higher up the river than San Jose. The limited time which 
could be spent in this region made the trip scarcely more than a 
reconnoissance, which it is hoped at some later day to follow up 
with systematic exploration. 

Distribution of Ruins in Pueblo Viejo » 

In ancient times, when the vallej 7 was populated by a sedentary, 
agricultural race, aboriginal dwellings were thickly scattered over 
the plain between the left bank of the Gila and the Graham or Pina- 
leno mountains. These dwellings were high up on the neighboring 
foothills as well as in the level plain, adjoining the river. In 
places houses were clustered together, forming a village, but the 
majority were isolated, dotting the whole valley. A compact, 
communal town of the pueblo type, such as is met north of the 
Apache reservation, was not found, and even when the population 
was concentrated the villages were composed of many clusters of 
small houses, separated from each other. As a rule, however, in 
such a cluster one central structure was much larger than the 
remainder. This centrally placed building, which is shown in the 
plan of the Buena Vista ruin (plate lxvi), resembles a type common 
in the Gila, Salado, and Verde valleys, where we find a central house 
surrounded by many mounds, indicating that a suburban population 
was settled about it. 

The majority of the clusters of mounds which were examined were 
situated in the plain not far from the river. This choice was evi- 
dently advantageous for an agricultural life, and the want of com- 
pactness in the houses would seem to indicate that the farmers had 
not yet been harried and driven to seek shelter from marauding nomad 
tribes in walled pueblos. 

fewkes] PUEBLO VIEJO KUINS 17 1 

Epley's Ruin 

This is the largest ruin in the vicinity of Solomonville, and lies on 
the outskirts of the town, on the road to San Jose. From its position 
it was the most convenient to study, and considerable work was done 
in the mounds which compose it. The majority of the mounds had, 
however, been leveled to the surface of the plain by Mr Epley, and 
as the place is a favorite quarry for adobe makers, their excavations 
have destroved most of the ancient walls. 

Just back of the Epley farmhouse there still existed (1897) one 
of the tallest mounds, which had been partially excavated by Mi- 
Adams. The author's party continued his work, but discovered 
nothing of interest save the Avails of rooms, all of which were of 
great thickness. From the size and position of the cluster the author 
concluded that it was the remains of the central building or citadel 
of the group. 

The smaller mounds which dotted the farm around it were traced 
almost to the river bank. The remains of house walls could be dis- 
covered in most of these, but excavations in the majority of the rooms 
developed very little of archeological worth. A few large ollas made 
of rough ware were taken from the mounds at the eastern end of 
the farm, but they were all broken. One or two slipper-shaped jars 
and food bowls of decorated ware were dug from the same rooms. 
Perhaps the most important objects from Epley's ruin were the 
skeletons of two infants, buried in the floor, accompanied by mortuary 
'bowls and small vases. 

A considerable number of whole bowls and vases were offered for 
sale by persons, mainly Mexicans, living in the neighborhood. It 
was reported that these had been taken from Epley's ruin by the 
adobe makers, and there is no doubt that such was the case. 

While the author was at the ruin a party of these laborers unearthed 
from the level land, a hundred yards east of Epley's house, a deco- 
rated vase (plate lxviii) filled with burnt human bones, which were 
secured and added to the collections. 

It was customary, before the burial of these cinerary urns, to cover 
the orifice with a circular burnt-clay disk, which was carefully luted 
in place with adobe. These urns were deposited not far from the 
pyral mounds, on which the cremation occurred, and were buried only 
a few feet below the surface of the ground. The adobe diggers 
reported that they always found a number of these ollas in close 
approximation, and that burnt bones were generally found within 


The best preserved of all the mounds in the Pueblo Viejo which 
were visited is situated at Buena Vista, a few miles east and north of 
San Jose, and is probably the ruin which gave the name to the 


whole valley; San Jose being sometimes called San Jose de Pneblo 
Viejo. The ruin of Buena Vista is typical of those lower down the 
river — of the mounds less disturbed ln T the farmer. Indeed, it is 
probably in about the condition in which all the ruins were when 
Emory passed through the valley. 

The site of the cluster of mounds of Buena Vista is a high bluff, at 
the base of which, on one side, flows the Gila river. A few modern 
adobe houses, inhabited by Mexicans, have been built ou the bluff, 
and some of the ancient walls have apparently been utilized in these 
modern structures. The largest and most conspicuous ancient build- 
ing is an irregular stone structure which is situated somewhat back 
from the edge of the bluff, and is now used for a corral. The walls 
which composed it have tumbled down, but enough remains to indi- 
cate its ancient form. Apparently it was formed of many rooms 
which were built about a central plaza; stones were extensively used 
in its construction. 

Surrounding this larger stone inclosure there lie at intervals low 
mounds, some of which betray evidences of rooms, while others are 
simply ash heaps. Two large circular depressions, a few hundred 
feet from the central building, are conspicuous. The limits of the 
cluster of mounds which compose Buena Vista could not' easily be 
determined, and probably no two persons would agree upon their 
extent. The more prominent, however, are sketched in the accom- 
panying plan (plate lxvi). 

It would hardly be consistent to call this cluster of mounds the 
ruins of a pueblo, as we ordinarily understand the word. They lack 
compactness and mutual dependence. The houses, save the large 
central building, are more like farm houses, or isolated buildings, of 
one story, with a few rooms, inhabited by a single family. They may 
better be known as rancherias, which have been arranged in a cluster 
for certain mutual advantages. Among these was probably nearness 
to a central house which might serve as a place of refuge, or, pos- 
sibly, for ceremony. The vicinity to the large circular depressions 
in the ground, which may be interpreted as reservoirs, was also a 
decided advantage. The presence of small mounds of ashes near the 
larger mounds containing remains of house walls would seem to indi- 
cate that each family had an individual burning place for its pottery. 
Possibby the dead were cremated on these mounds, which accounts 
for the absence of cemeteries, and for the ollas with calcined human 
bones sometimes found buried in them. 

Architecturally there is very little likeness between this central 
large stone inclosure or house with many rooms and Casa Grande, the 
best-known building of the Casa Grande group. This difference is 
in part due to the character of the building material, but more to the 
plan of the building itself. a The large central stone. structure of 

a From Mindeleffs valuable description of the Casa Grande groupof ruins it appears that Casa 
Grande was neither central nor the largest structure in the cluster of buildings. 


Buena Vista is more like those north of the White mountains and 
resembles closely the rectangular ruin at Pinedale. We have in the 
Buena Vista ruin resemblances to both the lower Gila ruins and 
those of the southern tributaries of the upper Little Colorado. 

Thus far in his archeological studies the author has failed to find 
in the belt of Arizona ruins from Sikyatki south to the Gila any 
rooms which he can positively identify as kivas or ceremonial cham- 
bers. As is well known, however, each of the modern Hopi pueblos 
ha* one or more of these rooms, though some of the important secret 
ceremonies in the modern Hopi pueblos are performed not in special 
kivas, but in the oldest homes of the clans. 

There was no room found in the Pueblo Viejo ruins which could be 
called a special ceremonial room, and in the large ruins at Chevlon, 
Homolobi, and Chaves pass no undoubted kivas were found. The 
room described in the preceding account of Four-mile ruin may, how- 
ever, be regarded as a ceremonial chamber. The kiva, as we now 
find it in Tusayan, is a late innovation, and was probably introduced 
from the eastern pueblos. Its existence in Four-mile ruin may be 
accounted for by the position of this ruin. 

Other Ruins 

Many objects of pottery have been dug up near the San Jose 
settlement, and there are one or two mounds near by indicative of 
ancient dwellings. If there ever was a large cluster of mounds on the 
present site of the town, they have been wholly obliterated by its 

There is a mound of some size on the right bank of the irrigating 
ditch, just as one enters the town from Solomonville, but one side of 
it has been worn away by freshets from the San Simon. It serves as 
a protection for the neighboring farm, which lies between it and the 
river, and on that account the owner refused to allow it to be dug 
away. A few days' labor at this ruin would bring to light objects of 
archeological value, for a beautiful vase rewarded an hour's superfi- 
cial scratching of the exposed bank. One of the finest ollas obtained 
from the Pueblo Viejo was purchased from a San Jose man, who dug it 
out of this mound while working at Buena Vista. As charred human 
bones Avere found in it, this vase, figured in plate lxix c, is regarded 
as a cinerary urn. 

There were formerly several mounds indicating ruins near Thacher, 
but these have been mostly leveled and can not now be traced. A 
number of mounds are still visible at Mr D. Olney's ranch, and lines 
of st oik.'s, the foundations of ancient walls, can still be traced in the 
road in front of Mr Lem Place's house. The large mounds on Mr 
Peter Anderson's farm have been destroyed, and there are many 
Others near it which have met the same fate. It may be said that in 
ancient times the houses of the aborigines dotted the valley through- 


out its entire length, from Buena Vista to Pima, and the indications 
are that the population was larger and had a greater number of acres 
of land under cultivation than at the present time. 

In a valley which was so densely populated we should expect to 
find a large number of antiquities, stone implements, ollas, and other 
forms of pottery. From all that can be learned comparatively few 
specimens have been dug out of the ground, although there are sev- 
eral private collections of some size. Different farmers have told the 
author of plowing off the necks of rows of buried vessels, and work- 
men on the irrigating ditches report finding pottery in abundance in 
several places far from mounds; but the large majority of relics are 
still under ground, and probably will remain there for years to come, 
now that the fields above them are cultivated. The only collections 
of any size which have found their way to public institutions, and 
are therefore available for study, are one of a few specimens in Tucson 
and that which was brought back to the National Museum. 

Changes in the Valley Since 1847 

Great changes have been wrought in the appearance of Pueblo Vie jo 
since Emory passed through it in 1846, for if any white man lived 
there at that time he says nothing about him. There were evidently 
no settlements, for he wrote: 

Everywhere there were marks of flowing water, yet vegetation was so scarce 
and crisp that it would be difficult to imagine a drop of water had fallen since last 
winter. . . . The dust was knee-deep in the rear of our trail; the soil appeared 
good, but for whole acres not a sign of vegetation was to be seen. Grass was at 
long intervals, and, when found, burnt to a cinder. 

In a prophetic way he added : 

The whole plain, from 3 to 6 miles wide, is within the level of the Gila, and 
might easily be irrigated, as it no doubt was by the tenants of these ruined 
houses. « 

Ancient mounds, in much the same condition as those in Pueblo 
Viejo formerly were in, still remain in the long stretch of country 
between Geronimo and Dudleyville, across the southwestern corner 
of the Apache reservation, wherever there are plains along the Gila, 
but white settlers have worked marvels in other parts of the valley, 
which may now be said, using a familiar simile, to "blossom as the 
rose." At present Pueblo Viejo, from Buena Vista to Pima, which 
towns mark the limits of the author's acquaintance with it, is one 
succession of cultivated farms of corn, alfalfa, 'and melons, a garden 
of Arizona in which any crop can be raised. 

It seems incredible that in fifty years such great changes should 
have taken place, yet it was to be expected, for in prehistoric times 

a Notes of a Military Reconnoissance, Washington, 1848, p. 68. 

fewkes] PUEBLO VIE JO RUINS 175 

Pueblo Vie jo was a garden spot, and there is every reason to believe 
that when it was inhabited by aboriginal farmers more acres of its 
land were under cultivation than at present. 

Former Population of the Valley 

If we judge from the number of ruins, the capacity of ancient reser- 
voirs, and the size of irrigating ditches, the extent of the terraced 
gardens, and other evidences of aboriginal agriculture, Pueblo Viejo 
was formerly densely populated. To be sure, there is no proof that 
all the ancient buildings were simultaneously inhabited, and, on the 
other hand, there is no reason to suppose that they were not. 

The aboriginal population was not huddled into a few beehive 
pueblos for protection, but was spread over the plains in small 
rancherias, or farming hamlets, dotting the valley from one end to 
the other. The evidences of the large ancient population are, how- 
ever, rapidly disappearing, and in a few years will have completely 
vanished. \ 

Cremation of the Dead 

There were apparently two methods of disposing of the dead prac- 
ticed by the ancient people of the Pueblo Viejo ruins, viz, house- 
burial and cremation. 

Evidences of the former method were found at Epley's ruin and at 
Buena Vista, and the same are reported from the ruins near Thacher 
and elsewhere. The skeletons found in house-burials at Epley's ruin 
were mostly those of infants, and were accompanied with mortuary 
food vessels and bowls, generally rude ware. It was also common to 
find metates in the neighborhood of such interments in such positions 
as to indicate that they were placed there by design. 

Evidences of cremation were common, consisting of calcined human 
bones in mortuary ollas, with ashes, evidently of bones, buried on 
certain low mounds adjoining the houses. It was apparently the 
ancient custom to burn the dead on certain pyral mounds and then 
to gather up the remains of the burnt bones and deposit them in small 
rudely decorated vases. A circular disk made of pottery was luted 
to the orifice of these vases and the whole was buried in an upright 
position near the edge of the mound upon which the burning took 
place. In its neighborhood there were also placed jars or other 
mortuary objects, as in the case of intramural interments. 

This method of disposing of the dead is similar to that adopted by 
the ancient people of the great ruins of the Gila-Salado region, add- 
ing one more indication of a close resemblance between the ancient 
inhabitants of the Pueblo Viejo and those lower down the Gila river. 

There survive among aboriginal people of the Gila-Salt valley two 
distinct forms of disposal of the dead, burial and cremation. The 


fact that some of the tribes in this region burned their dead and that 
others did not was mentioned by historians in the middle of the fif- 
teenth century, and it would seem possible that here we have evi- 
dences of two distinct stocks in the vallev. These two stocks had 
partially consolidated, forming a people which built the great houses. 
Certain clans of the compound stock, like their ancestors, cremated 
the dead; others interred their deceased. The custom of burning the 
dead does not appear to have been carried into the Little Colorado 

In this connection, statements of Castaneda that the Cibolans 
burned their dead is instructive. The author has not, however, dis- 
covered north of the Mogollones any archeological evidences of 
cremation, and is unaware of any well-authenticated statement that 
they have been found in any Zuni ruin. The suggestion that the 
present Zuiiians in mortuary customs perform certain ceremonials 
which symbolize burning the dead has been given some weight, but 
this might be interpreted as a survival transmitted to modern times 
by clans who came from the south. Our knowledge of the nature of 
this reported Zuiii survival is very vague. 


The houses of Pueblo Vie jo are arranged somewhat differently 
from those of the Little Colorado and its tributaries. The tendency 
in the latter regions is toward consolidation, toward a close approx- 
imation into a communal pueblo, while the buildings in the Pueblo 
Viejo are more like rancherias or farm dwellings. Each of the 
houses was small, apparently inhabited by a single clan, and they 
were generally grouped in clusters, which may for convenience be 
styled villages. 

There is generally found in the midst of, or near, such a cluster of 
small houses, a larger building which occupied the relation of a 
citadel, or, possibly, a ceremonial room; it may be single or composed 
of several chambers. This feature can be well seen in the accompa- 
nying plan (plate lxvi) of the Buena Vista ruin, one of the least 
changed of those in the Pueblo Viejo. The existence of a central 
room with clusters of small houses near or about it reminds one of the 
Casa Grande group near Florence, Arizona. 

There is no region of the Southwest from which better examples of 
the influence of environment on architecture can be cited than in the 
Gila valley. The majority of houses in portions of the valley where 
stones are absent were built of adobe, while in the upper part of the 
river valley, where rock is more abundant, we find that the inhabi- 
tants utilized it as a building material. Thus, while adobe forms the 
greater part of the walls of Casa Grande, the great central room of 
Buena Vista was constructed of rock. 


In the majority of houses in Pueblo Viejo there were three build- 
ing materials employed in the Avails, namely, stones, adobe, and logs. 

River-worn stones arranged in rows are in many places all that 
remain of the ancient walls of rooms. It would seem that they for- 
merly served as foundations, and were sometimes inserted in the 
sides of the house, but in neither case were the3 r closely fitted 
together. They imparted a certain solidity to the walls, and, when 
used in foundations, prevented erosion at a weak point in its struc- 
ture. No attempt to dress these stones, or, indeed, to break them, 
was noticed, but they were laid together with clay — the main building 
material employed. 

There were many and conclusive evidences that logs were employed 
in the construction of the house walls. These logs were driven 
upright along the lines of the foundations at short intervals, and 
gave strength to the walls and support to a roof which covered 
the chamber. The spaces between them were filled in with stone 
and adobe. 

In the early accounts of the ancient habitations of Pueblo Viejo 
by Emory and Johnston mention is made of these logs, and many of 
them were still standing in place when the Army of the West 
passed through the valley in 1847. Old residents of San Jose say 
that when they first took up their abode in the place the upright 
logs in some of the Buena Vista house-clusters were still visible. 
Only a few now remain above ground, yet the bases of several were 
discovered by the author's excavations. 

The rapid disappearance of these logs can doubtless be partly 
explained by their use as fuel. For years the mines in the neighbor- 
hood employed laborers cutting firewood, and the large mesquite 
bushes were used for that purpose. No doubt the logs of the early 
buildings were among the first gathered by them. 

Terraced Gardens 

Students of Southwestern archeology are familiar with rows of 
stones marking off the surface of the land in rectangles of great reg- 
ular it v. Some of these lines of stones extend for several hundred 


feet. They occur on level mesa tops or on side hills, but there is 
rarely any broken pottery or other evidences of human habitation 
about them. Various interpretations have been advanced to account 
for these regular rows of stones. By some authorities they are sup- 
posed to be the remains of house walls, or foundations of the same, 
and as such the} 7 are commonly pointed out to the visitor. Minde- 
leff speaks of them as "bowlder sites," and describes many from the 
Verde valley. Similar bowlder sites are very abundant, especially 
on the sides of the mesa bounding Pueblo Viejo, in the San Simon 
valley; probably a correct interpretation of them in these localities 

22 eth— 04 12 


would equally well apply to other bowlder sites, as, for instance, those 
of the Verde valley. 

The arrangement and size, and absence of remains of human life near 
these lines of bowlders have led the author to abandon the commonly 
accepted theory that they have relationship to house walls, or, indeed, 
to habitations of any kind. The small size of the bowlders employed 
shows that they are not fortifications, and they should not be con- 
founded with trincheras or fortified hilltops so common in southern 
Arizona and northern Mexico. They may be regarded rather as the 
walls of terraced gardens, so placed as to divide different patches of 
cultivated soil, or to prevent this soil from being washed down to the 
plain below. 

Very extensive terraced gardens may be seen not far from San Jose, 
and all along the mesa near the Solomonville slaughterhouse. It 
would seem from their distribution that not only irrigation ditches 
watered the valley of Pueblo Vie jo, but also that water was in some 
way carried up the hillsides, so that land now barren was in ancient 
times cultivated by the people of this region. 

As no remains of rancherias were found near some of these ter- 
raced gardens, it is evident that the farmers who tilled them had to go 
a considerable distance from their homes to plant and harvest their 

The use of terraced gardens still survives among the modern Hopi 
Indians, and these structures may still be seen on their reservation, 
at Wipo and Kanelba on the East mesa, as well as on the Middle mesa 
and at Oraibi. The size of the gardens on the East mesa is much 
less at the present day than in former times; those which have been 
abandoned closely resemble the rectangles inclosed by lines of stones 
in the Verde and Gila valleys. 

Prehistoric Irrigation in Pueblo Viejo 

There are evidences that the ancient farmers of the Pueblo Viejo 
irrigated their farms, for remains of extensive aboriginal ditches can 
be seen at several points. These old canals are clearly visible in 
that part of the valley which is not at present cultivated, but traces 
of them have naturally disappeared before the plow of the white 
settler. The remains of large circular reservoirs can be readily traced 
near some of the house clusters of Buena Vista, and not far from 
Epley's ruin, where there is a reservoir from which was undoubtedly 
drawn the water supply of that neighborhood. At the time of the 
author's visit this reservoir was full of water, which was used on the 

The modern acequias, the San Jose and Montezuma ditches, follow 
in part of their courses the ancient canals, as the author has been 
informed by an old settler in Solomonville; and a section of a side 
canal at right angles to the Gila may still (1897) be traced near San 


There can hardly be a doubt that water was carried in large earthen 
vessels to some of the terraced gardens, the altitude of which above 
the water in the river would make irrigation otherwise impossible. The 
surface of the land near the banks of the stream is continually shift- 
ing, on account of erosion due to heavy freshets and overflow of the 
river banks. On this account many of the ancient canals have been 
filled with soil, or their banks washed down to the level of the sur- 
rounding plain. 

Pottery from Pueblo Viejo 

The pottery from the Pueblo Yiejo ruins is identical with that from 
lower down the Gila river, at Phoenix and Tempe. It differs very 
markedly from that of the White mountains. 


As a rule the Gila pottery is coarse, and the decoration is simple, 
consisting mostly of rectangular geometrical designs. It may be con- 
sidered under the following types : 1, undecorated rough ware; 2, deco- 
rated rough ware; 3, undecorated red ware; 4, decorated black and 
white ware; o, decorated gray ware. 

Undecorated Rough Ware 

The larger ollas found in excavated rooms are almost always made 
of a rough coiled or indented ware of coarsest manufacture. These 
were capacious enough to contain several gallons of water, and were 
apparently used for that purpose. The exteriors of many were black- 
ened with soot, as though they had been used for cooking, as is at 
present the custom among the Pueblo Indians. 

Most of the large specimens of this rough ware were broken, appar- 
ently by the falling of walls or other debris upon them. It may also 
be mentioned that they were almost universally found in houses, 
and that one contained the skeleton of an infant. 

Small rough-ware vessels also occur, broken or entire (see plate 
lxvii). The author has limited this group to those specimens of 
pottery of rough ware in which there is no shining black slip on the 
inner surface. No food vessels of rough ware were found, but all 
specimens of this form, of which there were many, had a polished 
black interior, and belong to the second group. 

Decorated Rough Ware 

In this group are placed those food vessels in which the interior is 
covered with a black slip, which reminds one of the modern ware of 
Santa Clara pueblo. As will be seen by consulting a plate showing 
this type (plate lxvii), there is some variation in the arrangement of 
the indentations and coiling in this ware, but no color decoration 
was attempted. Bowls of this kind are often rubbed smooth on the 
outer surface, but decoration by indentation or coiling is common. 


Undecorated Red Ware 

A number of pieces of pottery of bright red color, made of coarse 
paste, were found in the Pueblo Viejo ruins. These were smoothly 
polished on the exterior, but as a rule were not decorated. In gen- 
eral appearance they resemble the ware still made by the Papago 
Indians, and the} 7 were commonly large, narrow-mouthed vases. This 
kind of ware was found to be abundant in caves where sacrificial 
vessels were found. Disks made of it often cover cinerary vases. 

Decorated Black and White Ware 

The white ware with black decorations, generally in geometrical 
designs, was sparingly represented in the Pueblo Viejo ruins, which 
is in marked contrast to its prominence in cliff houses near the sources 
of the Gila in New Mexico. This ware is so rare in the vicinity of .San 
Jose that the author is inclined to regard it as intrusive in that region. 

None of the specimens found are at all comparable in the wealth 
of their ornamental designs with similar ware from ruins in the 
western part of Socorro county, New Mexico, a or in the cliff houses 
near the sources of the Gila and Salado rivers. ^ 

Decorated Gray Ware 

The characteristic decorated ware of the Gila valley and its tribu- 
taries is grayish and is decorated with red; a specimen is shown in 
plate lxviii. With the exception of a few sporadic specimens which 
have been transported to pueblos, now ruins, north of the White 
mountains, this ware has not been found in any valley except those 
of the Gila and its tributaries. 

TJiis pottery bears a smooth polish, is never glazed, and is generally 
decorated with geometrical figures: scrolls, terraces, stars, and key 
patterns. It assumes a great variety of shapes, and was apparently 
used in much the same way as is the yellow or red ware of northern 


The decoration of the pottery from this region is mainly in geometri- 
cal patterns, resembling that of the pottery from the Little Colorado 
basin. No specimen with picture writing was found, so that this source 
of information regarding the mythology of the owners is practically 
wanting. Even pictures of birds, so common north of the Mogol- 
lones, are absent. The ancient. people of the Pueblo Viejo had not 
carried pottery decoration beyond the geometrical stage, as far as can 
be judged from the specimens examined. Food bowls, almost identical 

a The beautiful collections made in this region by Mr H. Hales are now preserved in the 
National Museum, and are well worth study and description. For an account of the ruins 
near Tulerosa see N. Francis Duff, The Prehistoric Ruins of the Rio Tulerosa, in Bulletin of the 
American Geographical Society, v. 29, n. 3, 1897. 




in form with those of Tusayan, have their interiors decorated with 
rectangular patterns, sometimes with terrace figures, but rarely with 
spirals. Encircling bands are often broken at one point, forming 
' ' life lines, " and zigzag lines are not uncommon. Few specimens with 
external decorations Ave re found. Vases were generally decorated with 
the same simple geometrical patterns as were the food bowls, with no 
attempt to depict human or animal forms. It is unfortunate for 
the student of Gila valley 
ceramics that pictographic 
material is so scanty, as it 
shuts him out from most 
instructive data regarding 
ancient life in this part of 

A finelv made human head, 
forming the handle of a 
ladle, was dug out of an 
ash heap in the Buena Vista 
ruin. This was the only 
handle modeled in human form which was found, though objects of 
of this kind are said to be common lower down the Gila river. A 
like ornamentation is not a rare feature of ladle handles from the Little 
Colorado ruins. 

The broken encircling band, called the "line of life," occurs on 
many of the decorated pieces of pottery, and the H figure, so com- 
mon on the exterior of bowls north 
of the mountains, was found on a 
single fragment of pottery from 
Pueblo Vie jo. 

Fig. 110. Decorated slipper Jar from Pueblo Viejo 

(number 1775.33). 

There were a few specimens of 
clay slippers which were orna- 
mented on their upper side. One 
of these is shown in the accom- 
panying cut (figure 110), in which 
the design of the decoration can 
be partially seen. It is not im- 
probable that the " foot of an idol," 
mentioned by Emory and his offi- 
cers, was one of these slipper jars. 
Another specimen from the Pueblo Viejo, in which the likeness to an 
Indian moccasin is close, is shown in figure 111. The specimen is, 
however, much smaller than a human foot. 

The accompanying illustration, plate lxviii, shows two views of 
one of the most richly decorated vases from the Pueblo Viejo, and 
exhibits several of the commoner geometrical designs from the Gila 
ruins. This vase was dug up near San Jose, and was probably a cin- 

FiG. 111. Moccasin-shaped jar from Pueblo 




[ETH. ANN. 22 

erary olla, as it contained, when found, cinerated human bones. The 
ware is characteristic of the Gila, though a few straggling specimens 
of similar pottery have been found at Four-mile ruin near Snow- 
flake. It will be observed that the decoration of this vase is wholly in 
geometrical patterns, a common feature of all ornamented ware from 
the Pueblo Vie jo. Almost all geometrical forms are represented — 
spirals, bars, terraces, stars, and squares with dots. 

In plate lxix other forms of decorated ware from Pueblo Viejo 
are represented. Figure a shows a small saucer, with exterior and 


Fig. 112. Arrow polisher from Pueblo Viejo (number 177569). 

interior decoration of rectangular bands of black ; the margin is white. 
Figure b shows a small vase of typical Gila pottery, ornamented with 
zigzag red bands, which was excavated from the Buena Vista ruin. 
The vase c was dug out of the flat near the western mounds of Epley's 
ruin. It was found by Mexican laborers making adobes, and con- 
tained a calcined human skeleton. The external surface of this vase 
was smooth, and the decoration consisted of series of terraced figures, 
recalling those geometrical designs so prominent in all ancient pottery 
from Arizona. 




Stone Objects from Pueblo Viejo 


The ancient people of Pueblo Viejo were still in the stone age, and 
their implements were similar to those found elsewhere in the South- 
west. The stone hatchets are, as a rule, finely made, as is generally 
the case in the Gila and Salt river ruins. A considerable number 
were collected, some of which were among the finest known to the 
author. They were, however, identical with stone implements that 
have already been collected in other parts of Arizona. There was 
nothing strikingly peculiar in the arrow and spear points collected in 
this region. The stone axes were finely polished and very numer- 
ous. There were many hammer stones, 
pounders, rubbing stones, stone knives, 
and drills. 

Although nothing distinctive was noticed 
in the arrowheads, a fine collection of these 
implements made of volcanic glass, from 
the cave in the Nantacks hereafter dis- 
cussed, should be mentioned. A number 
of spherical stones, varying from the size 
of the fist to that of a large marble, were 
picked up on the surface of the mounds. 
Some of these may belong to a type of stone 
objects referred to in early accounts as 
being used by the people in warfare. 
They are thus mentioned by Castaneda: 
"Farther off was another large village, 
where we found in the court-yards a great 
number of stone balls of the size of a 
leather bag containing one arroba. They seem to have been cast 
with the aid of machines, and to have been employed in the destruc- 
tion of the village." What the nature of this machine was we are not 
told, but it was possibly a kind of sling. Problematic stone disks 
with depressed faces and circular forms are not rare. Simple si one 
disks of the same shape, but perforated, ma}^ once have been attached 
to drills. There were pottery disks which were supposed to have 
had a like use. 

Arrow polishers (figures 112, 113) or grooved stones for rubbing 
down wooden sticks occur in numbers. The depressions in some of 
these were so smooth that their efficiency in grinding must have been 
small, while in general shape and size they correspond so closely with 
those stones which are still used for that purpose that there can be 
little question as to their use. 

A large number of metates, or stones for grinding corn, were col- 
lected in the excavat ions at Kpley'sruin. These were made of several 
kinds of rock, the favorite being lava or malpais. Evidences of long 

Ftg. 113. Arrow polisher from 
Pueblo Viejo. 



[ETH. ANN. 22 

use can be seen in the deep depression which has been ground into 
them, even when the rock was very hard. 

The form of metate Avith three legs (figure 114) is more Mexican than 
any which was found north of the White mountains or in the Little 
Colorado valley, and is not unlike rude specimens from Central 

Small stones with a regular depression in one face were used for 

Fig. 114. Metate from Pueblo Vie jo (number 177471). 

grinding pigments, and the remnants of color were still found on 
them. Green, red. yellow, and white pigments such as are still used 
by the Pueblo Indians were found in many of the rooms. Oblong 
or oval stones, with a flat face, worn oh one pole, served as rubbing 

stones by which these 
substances were ground 
to powder. There was 
sometimes considerable 
skill shown in the way 
these stones were fash- 
ioned. They were some- 
times mushroom-shaped, 
with a circular disk and 
a slim handle. 

One of the finest 
wrought of all the mor- 
tars was purchased from 
a Mexican in Solomon- 
ville. It was elongated, 
trencher-shaped, with 
knobs at the extremities. 
The rock of which it was made, though very hard, was worked with 
considerable skill. The Mexican who sold it had used it for bruising 
vegetable substances. No doubt this is but a continuation of its use 
in prehistoric time, long before white men came into the country. 
The author saw a beautiful mortar a made of a green stone, which 

Fig. 115. Unknown stone object from Pueblo Viejo 
(number 177677). 

« Attempts to purchase this fine specimen failed. 




was said to have been taken out of the ruins near Solomonville. This 
was one of the finest paint mortars which he ever saw from the 

One of the most exceptional of stone objects from the Pueblo Viejo 
ruins is shown in the annexed cut (figure 115). It has a regular 
disk form, and is carefully worked from a lava stone. The form 
is that of a paint mortar. 


Early in the author's studies in the Pueblo Viejo, his attention was 
called to a stone slab shaped like the sole of a shoe (figure 116), to 
which it was compared by the Mexican who owned it. This object 

Fig. 116. Ceremonial stone blab from Pueblo Viejo (number 17~">75). 

was flat or slightly convex on one face, flat on the opposite, and had a 
shallow groove on the margin. The border on the flat side was orna- 
mented with a number of parallel scratches arranged in clusters. 

Later the author obtained other stones of the same shape and of 
about the same size; one of the most instructive was a specimen of 
irregularly rectangular form, with a bird's head carved on one edge, 
and the tail on the other ( figure 1 17). 

There is an interesting modification of the same class of objects 
in the collections of the National Museum — a circular stone slab of 
which the bod} 7 of a snake, with head and tail skillfully carved, forms 
the margin. These objects, which are not rare in the ruins of 1 lie 
Gila and Salt river valleys, are called ceremonial slabs, and were 
probably used in much the same way as are the stone slabs orna- 



[ETH. ANN. 22 

mented with designs which have been repeatedly described from 
Hopi altars. In fact, some of those now in nse distinctly resemble 
those from the Pueblo Viejo. 


It was interesting to find in the ruins of the Pueblo Viejo a number 
of obsidian nodules, and flakes of the same material. Fragments of 

Fig. 117. Ceremonial stone slab from Pueblo Viejo (number 177578). 

volcanic glass constantly occur in ruins north of the mountains, along 
the Little Colorado, and at Sikyatki. 

The fact that over a dozen specimens of quartz crystals were found 
shows that these objects were prized, and were probably used in cere- 
monials, as is the custom 
in modern pueblos. 

Perforated circular 
stones, varying in size 
from that of a silver dol- 
lar to several inches in 
diameter, were found, 
while similar disks made 
of pottery were common. 
Manj' of the latter were 
n ot perforated. T h e i r 
form suggested that the} 7 
were used in gaming. 
These disks occur along the whole length of the Gila river, and are 
also found more sparingly in ruins north of the Mogollon mountains. 
The ancient inhabitants of the Pueblo Viejo villages prized for 
ceremonial purposes stones of curious or strange shape, especially 
those which were botrvoidal. One of the most striking of these is 
shown in the accompanying cut (figure 118). It is typical of several 
which were found in the San Jose ruin, at Buena Vista, and near 

Fig. 118 Stone fetish from Pueblo Viejo. 

fewkes] RUINS VISITED IN 1897 187 

Shell Ornaments from Pueblo Viejo 

Several specimens, generally ornaments, carved out of Haliotis, 
Conus, or Pectunculus shells, were collected in the Pueblo Viejo 
ruins. The largest of these were armlets and wristlets. Shells cut in 
the form of animals are among the characteristic objects of the Gila- 
Salado ruins, and occur in all great collections from this region. We 
find various animals represented, as lizards, birds, rabbits, and 
snakes, as well as circles, crosses, rectangles, and the like. They are 
ordinarily perforated — generally at the eye, sometimes at the heart, 
of the animal represented. The universality of this perforation 
implies suspension, and the author believes that it will later be found 
that they were worn on the neck or body. It is not improbable that 
they were personal fetishes, possibly representing totems. 


There are many cliff houses and other ruins of aboriginal dwell- 
ings in a fair state of preservation along Bonita creek, 18 miles north 
of Solomonville. A very good account of these, written b} 7 Professor 
William Stone Devol, of Tucson, has been published, with a half-tone 
illustration, in the Graham County Bulletin. These remains would 
repay more extensive study and no doubt yield collections of consid- 
erable archeological value. These cliff houses resemble in general 
character those near Silver City, New Mexico. 


There are many ruins, mostly small, on the mesas and foot hills 
of Mount Graham, having the same general character as those 
lower down the valley, near the river. The fragments of pottery 
strewn about upon them are identical with those from the mounds 
of the plain, and there is reason to believe they were made by the 
same people. While these ruins occur at several places on the mesa 
and hills at the base of Mount Graham, that at the place called Cie- 
nega is one of the largest, but it does not differ radically from those 
of the banks of the Gila. 


The use of caves for ceremonial purposes was a feature in the life 
of the ancient people of Pueblo Viejo. The mountains near the 
Pueblo Viejo have many caves suitable for this purpose. They occur 
in limited number near the modern Ilopi pueblos and elsewhere north 
of the Mogollones. 

The author visited one of these sacrificial caves in the Graham 
mountains, and found many evidences of its former ceremonial use. 
There were bushels of prayer sticks on the floor, and a few fragments 
of basketry, but no pottery or earthenware rewarded the search. 


The fragments of basketry were made with a technique similar to 
that of the basket plaques of the Middle mesa. The prayer sticks 
were painted red at their extremities, and were, as a rule, about the 
size of a penholder. This cave, called Adams' cave, has been rarely 
visited since its discovery by Mr B. B. Adams, of Solomonville, but 
will well repay a visit by an archeologist. There is little doubt that 
there are other similar caves on the northern side of the Graham 
mountains which have not been entered by white men. 


In the broken, almost inaccessible, country north of Pueblo Viejo, 
there are many caves, some of which are quite extensive. The larger 
and more open were utilized by ancient builders in the construction 
of cliff houses. Many caves in this region have narrow entrances into 
passages which extend with many ramifications far into the bowels of 
the earth. The nature of the objects found in most of them shows 
that the caves were not inhabited, but were resorted to for purposes 

of prayer and sacrifice by a sedentary people 
akin to that which has left so many ruined 
houses in the Southwest. 

A few years ago some young men from 
Pima, a settlement in the Pueblo Viejo, ex- 
plored one of the caves in this region, and 
obtained from it a collection of some size and 
Pig. 119. indented bowl from considerable archeological interest. 
? Ca l e i~™ Nantacks The attention of the author was called to 

(number 1< <458). 

this collection early in the summer of 1807, 
and in September he visited Pima, and secured many of the objects 
for the National Museum. 

The collection has been divided by those who obtained it into four 
parts, one of which had been sent to Utah : another was owned by a 
man who did not care to sell. The other two parts, numbering almost 
100 specimens, were purchased. They contain all the typical forms 
of the other two, and a few specimens which were unique. 

All the specimens were small votive offerings, but those who had 
entered the cave declared that the} 7 left behind all fragments, of which 
there were many, so that we may suppose that there were larger 
vessels thus abandoned. The specimens were, as a rule, rough ware 
pottery (see figure 119), smooth, undecorated red ware, turquoises, 
arrowheads, fragments of marine shells, and white pigments. 

A large number of clay disks occur in the collections. These were 
not unlike similar formed objects which were found at Epley's ruin, 
and probably were originally luted to the orifice of the small vases 
in much the same way. 

There were globular vases (see figure 120), one of which had two 
perforated tubercles, one on each side of a small orifice. Tlmse 




Fig . 120. Small amphora from 
a cave in the Nantacks (num- 
ber 177463). 

remind one of those vessels in which sacred water is carried in cere- 
monies among the Pueblos. 

Another small "lobular vase had the whole exterior covered with 
indented tubercles, not perforated bnt evidently ornamental. This 
type has been found in some of the Little 
Colorado ruins. A long tube with similar 
tubercles over its surface, made of rough pot- 
tery, may have been an ancient pipe or cloud- 
blower. Neither of these objects had designs 
painted upon them. 

From the great quantity of turquoise beads 
and obsidian arrow-points it would appear 
that large numbers of these objects were scat- 
tered over the floor of the cave. As the col- 
lectors exercised no special care to gather everything which they saw, 
no doubt the quantity of these objects could be much increased by 
a reexamination of the cave. 


Pottery objects in the form of human beings are manufactured in 

some of the modern pueblos, and 
these grotesque figures may be pur- 
chased in traders' shops where mod- 
ern Pueblo pottery is sold. An ex- 
amination of large collections of an- 
cient pueblo pottery from northern 
and central Arizona has failed to re- 
veal a single specimen of a vase made 
in the human form. This, however, 
is not true of pottery from all parts 
of the pueblo area. The ancient peo- 
ple of southern Arizona manufac- 
tured human effigies in clay, the 
typical forms of which have not, so 
far as is known, been described. The 
particular interest attached to the 
vases here described, which justifies calling them into prominence by 
special mention, is due to the rarity of this type in ancient pueblo 
collections, its reappearance in certain vases from Arkansas, and its 
common occurrence in the northern States of old .Mexico. 

The accompanying illustration (figure 121) shows one of these vases 
from the cave in the Nantacks mentioned above. It is made of 
coarse material and has a rough exterior, with patches of a calcareous 
deposit on the surface. This deposit of lime is found in greateror 
less amount on most of the specimens from this cave, and was depos- 
ited on them by water charged with lime percolating from the rocks 

Fig. 121. Human effigy vase from a cave 
in the Nantacks (number 177519). 


in which the cave was formed. Attempts to rub off this film are evi- 
dent in some places ; but elsewhere, as under the right eye, consider- 
able patches remain, probably concealing symbols on the right cheek. 

The head is marked off by a constriction representing the neck, 
and the eyes, nose, mouth, chin, and ears are well represented. As is 
generally the case with idols of stone, wood, or clay from the pueblo 
region, the details of the head are better represented than those of 
the body or limbs. 

No attempt was made in this vase to represent the legs, and the 
arms are simply irregular ridges, one on each side of the body. The 
shape of the body is irregularty globular; the base is flat. The vase 
is of about uniform thickness, the outlines of its cavity conforming 
in a very general way with the elevations and depressions of the 
outer surface. 

The author supposes that this vase was filled with votive offerings 
when it was placed in the cave, and that in course of time its contents 
were washed out. The nature of the offerings may be conjectured 
from the fragments of shells, turquoises, and other objects reported 
as strewn about the floor of the cavern. 

The short parallel lines painted with white pigment under the eyes 
are worthy of a passing notice. These are the only symbols on the 
face, and consist of a few short lines extending downward from the 
lower eyelids. If the reader will examine the collection of Zuhi dolls 
which are exhibited in the " Pottery Court" of the National Museum, 
now installed, he will find one labeled Zuni Hehea katcina, a which 
has the same markings on the cheeks as has the effigy vase from 
the Nan tacks. 

It is instructive to note the similarities of this effigy vase with those 
from Casas Grandes, Mexico, and from Central America, which are so 
close that the vase might readily be mistaken for an illustration of a 
type from northern Mexico or even Central America. 

It appears that while this vase has a form unknown in collections 
of ancient pottery from ruins along the Little Colorado and its tribu- 
taries, it is not unique in those from the Gila-Salado watershed. 

The lesson taught by the presence of this effigy vase in the Nantacks 
and the Gila-Salado basin and the absence of similar forms north of 
the Mogollones may be summed up in two words, "Mexican influence." 
The distribution of this form of Mexican ceramics did not cease at 
what is now the southern frontier of Arizona, but extended to ruins 
along the Gila valley and its tributaries high up into the highlands 
to the north, where these streams rise. As far as is known, this was 
the most northern extension of this particular form of ceramic tech- 
nic in Arizona. Southward from this locality the relative number of 

"Hehea katcina is a Hopi nanie, and the doll representing this person at Walpi has not the 
same markings on the face as the above. The Hopi variant has parallel zigzag lines above both 
eyes and on the cheeks. The name given above is that by which the Zuni doll is known to the 




human effigy vases increases, unil they become very common in Chiapas 
and Oaxaca. But, it may be urged, wh} 7 is it necessary to interpret 
this form as due to Mexican influence? The advocate of the inde- 
pendent evolution of technology will doubtless say that the manufac- 
ture of a human effigy vase is no great trick and had been evolved 
independently again and again in different regions of aboriginal 
United States. Some clever potter of the Nantacks, it may be said, 
invented this form. Why, it might be asked in reply, did not the 
potters north of the Mogollones also invent the same form? for they 
were equally skilled, and their ceramic ware was more variegated 
and elaborate. What explanation is offered on the theory of inde- 
pendent invention of the increase in the relative number of effigy 
vases as we go south? 

It seems probable that the presence of human effigy vases in south- 
ern Arizona and their absence in 
the northern part of the Territory 
is in harmony with a theory of 
the influence of Mexican art in 
the former region. While rec- 
ognizing the potency of this in- 
fluence in southern Arizona, we 
are not necessarily called upon 
to accept a connection among all 
potters who have made human 
effigy vases, or even between 
those of ancient Arkansas' and 
Chihuahua, whose effigy prod- 
ucts have some similarity. 

There are many like ceramic 
forms and decorations among 
different people, invented inde- 
pendently, and there is no reason 
to doubt that human effigies in 
the form of vases were so in- 
vented in several well-known 

instances. There are also cases where identity in form and symbol 
can better be explained by barter. Possibly the effigy vase described 
above belongs to the latter category. It would be premature to 
build conclusions on a single specimen, and more information regard- 
ing the distribution of ancient human effigy vases in the Southwest is 
desired. These vases have not yet been found in Arizona north of 
the White mountains, but they are represented from several local- 
ities in the south. The question awaiting answer is, What is their 
northernmost extension? 

An effigy vase found near San Jose (figure 122) is instructive as 
recalling a kind of pottery common in the northern Mexican states. 

Fig. 122. Effigy vase from Pueblo Vie jo 
(number 177332). 


This piece was brought for sale by a workman, who declared that he 
had dug it up at San Jose. The author was at first inclined to believe 
that it was not found in Pueblo Viejo valley, but critical examination 
of the wire convinced him that the testimony of the man who brought 
it could be trusted. It is made of coarse red ware, like other vessels 
from this locality, and is undecorated. It is shaped like a dumbbell, 
and the two parts are of unequal diameter. The remarkable thing 
about this vase is the human nose and ears, in relief, reproduced sev- 
eral times on its sides. This would hardly be worth}^ of special men- 
tion were we considering the pottery of old Mexico or of some other 
parts of the United States, but like the effigy vase above mentioned, 
from the caves of the Nantacks, it is exceptional in the pueblo region. 

One of the best specimens of clay effigies was found in an ash heap 
at the Buena Vista ruin. It was evidently a handle of a dipper or 
saucer, and was well made and well proportioned. 

The frequency with which these effigy ceramic objects occur as we 
go south is, as has been stated, highly suggestive. Unknown in the 
ancient ruins of northern Arizona, they are not rare in the Gila valley 
and its tributaries, and their number greatly increases when we pass 
the boundary line into the Mexican states of Sonora and Chihuahua. 
This is undoubtedly an advance in pottery manufacture, and, with 
this advance a corresponding decline in the decoration of vases with 
paintings is to be expected. 


In plate lxx the author has tried to plot the distribution and rela- 
tive abundance of different colored pottery in the ruins studied by 
him in the years 1895, 1896, and 1897. A normal line is represented 
on one side and the relative amount of each kind of colored pottery is 
indicated by abscissas from that norm, arising from a point repre- 
senting the latitude of each ruin. In order to determine the proper 
percent of the kind of pottery in each ruin, the number of pieces 
obtained was counted, and the proportions of those referred to 
different colors were reduced to decimal fractions. In the case of red 
ware this was onty an approximation, for the limit of this type was 
hard to determine. 

Certain general laws may be deduced from a study of this map. 
Black and white ware, which is so prominent a feature of cliif-house 
pottery, has a limited distribution in all the ancient pueblos south 
of Tusayan proper. Its proportion increases in the Kintiel zone. 

Yellow ware is the characteristic pottery of Tusayan and is limited 
to the ruins near the inhabited Hopi villages. It is not represented 
at Kintiel. 

• Red ware is characteristic of the Little Colorado. Red, white, and 
black ware is not found in the north or south, but only near the 
Little Colorado and its tributaries. 


A brown pottery, with black decoration and red bands, is charac- 
teristic of the Gila valley ruins. 

Study of the material collected in 1897 suggests the conclusion that 
the higher we ascend the Little Colorado river the greater are the 
differences between the archeological objects found on its banks and 
those of the ancient Hopis who lived at Homolobi. These differences 
seem not to diminish as we ascend the Zuni river and other tributaries 
of the Little Colorado in the Zuhi reservation. At Four-mile ruin we 
find both Zuni and Hopi characteristics in ancient pottery, and no 
doubt some of the people of this pueblo were akin to the ancient Zuni 

There was probably not so close a likeness between the ancient 
people of Pueblo Viejo and those of modern Tusayan as between those 
of the Verde and Tonto villages and the Hopis, although there is a 
resemblance among all the ruins of the Gila valley and its tributaries. 
As a general rule, the culture of prehistoric peoples dwelling along 
the banks of a river has a marked uniformity, while that of those 
separated by mountain ranges is more varied. There is therefore a 
general likeness between the art products of the Gila valley and all 
its tributaries, and those of the Little Colorado are similar, but the 
archeology of the two drainage areas differs considerably. 


The Spanish word "pueblo" has come to be used in ethnology 
with a special meaning, and is now applied to a certain kind of 
Indian dwelling. While the Spanish explorers applied the term to any 
large cluster of houses it is well to limit it, as is now generally cus- 
tomary, to a communal village in compact form, with the different 
rooms adjoining. In this restricted meaning the clusters of houses 
in the Pueblo Viejo are not pueblos, but are better called composite 

A pueblo, then, may be regarded as a collection of rancherias the 
component houses of which have become so approximated that the}' 
adjoin, forming a compact village. Each clan has its own rooms and 
has no rights in others, though the walls may adjoin. 

We have a very good illustration of a communal form of archi- 
tecture in early Mormon settlements, as Brigham and Sunset, now 
in ruins near Homolobi. When these towns were built they were 
palisaded, and all the different families were protected by an inclosing 
wall. The houses joined, inclosing a central open space, much as in a 
small pueblo. Had there been no danger from Apaches or other 
predatory Indians, these Mormon families would probably have set- 
tled on separate farms, but it must also be borne in mind that there 
was community of life among the inhabitants which does not exist in 
Pueblo settlements. Each clan in the latter is independent; all fam- 
ilies in the Mormon towns mentioned had common property. This 

'22 ETH— 04 13 


community of life no doubt explains in part the pueblo-like character 
of the Mormon settlements, but mutual defense was an important 
factor in the determination of the form of their villages. 

The pueblo, therefore, as we find it today, is a survival of con- 
solidated cliff houses, cavate villages, or rectangular and circular 
towns of the plain, which have assumed their form for the sake of 
defense. But these forms are secondary; in localities and at times 
when defense was not necessary the aboriginal farmers erected more 
or less isolated dwellings or ranches, each with few rooms and with 
accommodations for one clan. 

In verv ancient times the inhabitants of the Gila were scattered 
over the land, or their homes were clustered together, but were not 
united in a compact form with adjoining walls. Even then, however, 
they had certain common houses for defense or religious purposes, of 
which Casa Grande is a good example. 

As the clans moved into exposed regions in which they were raided 
by hostiles they naturally built their houses in pueblos or forms best 
calculated for defense. 

It is interesting to note that when this pressure of necessity for 
defense was removed the former distribution of small farmhouses 
over the land returned. When the clan was no longer forced to 
huddle under the same roof with its neighbor, it returned to the 
isolated rancheria. In this way large pueblos have disintegrated, 
first into summer farming villages, later into individual farmhouses. 
Thus, a law of the struggle for existence and survival of the fittest 
can be educed to account for pueblo architecture in the Southwest. 
There is nothing in an arid environment to lead agriculturists to hud- 
dle into pueblos, and it was not until nomadic robbers forced them to 
do so that they adopted this form of life. 

The semi-deserts of the Southwest are not valuable lands for agri- 
culture, and yet the aboriginal people of this region were preemi- 
nently farmers. This is explained by the fact that it was impossible 
for hunters to remain in that culture stage, for there was no game; 
it was alike impossible to be fishermen, because there were no fishes. 
The people were forced by pressure of climatic conditions either to 
become farmers or to perish. In more fertile lands, where game was 
abundant, there roamed nomadic hunters with whom they were unable 
to successfully contend. Thus in an arid desert land the individual 
farmer became secure in his poverty from his warlike fellow-man. 
When, by his industry, he gathered property beyond his immediate 
needs, the nomads sought him out to despoil him of his possessions. 
To meet these attacks he joined his neighbors, building his houses in 
clusters, which, for additional protection, were finally consolidated 
into a pueblo form. As the enemies grew stronger the size of pueblos 
increased by consolidation. The form which the builders adopted 
was that best fitted fo r ' mutual protection. It has always been so 


with agricultural man when pressed by his foes, and on this account 
a cliff-building stage of culture is limited to no race or country. Its 
existence is purely a geological question. 

The Southwest is thus full of ruins of former abodes of farmers, 
some of which Avere inhabited by a single clan, others by several 
clans. Each has had its own history or its own episode in the 
general history of the struggle of nomadic robber and sedentary agri- 
culturist. Aimlessly to himself, perhaps, but in obedience to a law 
of development, man has drifted from place to place to escape his 
enemies, until he has been molded into the peculiar culture which we 
call Pueblo. This culture is a highly specialized form, and is the 
direct outgrowth of the peculiar climatic conditions of the Southwest. 


































r ^ 
















' "jm 




























































































































































































































































fePSVvA | 
































































■ ■ - 


















































































k i 

»• I'M 1 -' '< 


#1 , | 




vim w-n 'k x ■■■' Iw* x - vr* 

































































1 -* 





* • 



















_l I- 

co o 


















"«<» ^%\\. i 









a(NUMBER 157 558, HEIGHT Q'/ z INCH ES 






(NUMBER 157361, DIAMETER 9 3 / 8 INCHES ) 























































~- ■: , ' '>-" 


b{ NUMBER 157510, DIAMETER 5 3 / 4 INCHES 







• •'.NUMBER 157143. DIAMETER 4>S INCHES') 











» % * 











(NUMBER 157 1 19, DIAMETER 5%. INCH ES ) 









( NUMBER 157240. DIAMETER 7 3 / 4 INCHES ) 






a (NUMBER 177219, Dl AM ETER 9^2 INCHES) 











6 (NUMBER 177 356, DIAMETER 10 7 4 INCHES) 













R 158071, ACTUAL - 


























> V 1 P^ » 

















































a (NUMBER 157 142, Dl AM ETER 7^2 INCHES) 

















b{ NUMBER 157818, DIAMETER 9 \ I NCHES » 























A, altar; B, barn; C, cemetary ; G, gateway; H, modern houses; M, mounds; S, spring; T, trees. 




























F, fireplace, H, holes; P, posts; S, shrine. 























a NUMBER 177 162, Dl AM ETER 5% I NCHES ) 

6 I NUMBER 177 147, D I AM ETER 9% INCH ES > 


NUMBER 177 327, DIAMETER 10 V 8 IN 







































...jar .?>■< 









#V£ * ^s 

-"-"•'•V ^ -•• S«»™|i«"A 

•V-m. .■*-.-. >\ ^- .-4 





*>? -til"/ 






a, modern adobe buildings; C. H., central house ; D, depression; E, excavations ; m, mounds; R, reservoir; 

T, trees. 



- .•—SACi. 

a (NUMBER 177 544, DIAMETER 6 3 A. INCHES 

b (NUMBER 177 566, DIAMETER 9/ 2 INCHES ) 

* .**. >■)* y ■%.- ^ 

*"' 1 "'WTijii- _ii ^ _i _l _i i 










a (NUMBER 177536, DIAMETER 4/ 2 INCHES) 

b( NUMBER 177520, DIAMETER 3 [ /z INCHES) 












Prefatory note 203 

Initial series of Mayan inscriptions 203 

Secondary numeral series of the Quirigua inscriptions 226 

Maya chronological system 234 

The Cakchiquel calendar 275 

Maya method of calculation 282 

Signification of the numeral series 290 

Inscription at Xcalumkin, Yucatan 299 

Inscription on Stela 6, Copan 301 

The nephrite stone of the Leyden Museum 302 

Calendar and number tables 303 



Plate LXXI. Inscription on the west side of Stela F , Quirigua. Maudslay, 

part 12, plate XL 204 

LXXII. Inscription on the west side of Stela E, Quirigua. Maudslay, 

part 12, plate xxxi 208 

LXXIII. Inscription on the east side of Stela F, Qnirigua. Maudslay. 

part 12, plate XL 210 

LXXIV. Inscription on the back of Stela J, Quirigua. Maudslay. 

part 12, plate xlvi 214 

LXXV. Inscription on the east side of Stela A. Quirigua. Maudslay, 

part 11, plate vn 216 

LXXVI. Part of inscription on Stela D, Copan. Maudslay, part 2. 

plate XL vin 222 

LXXVII. Part of inscription on Stela D, Copan. Maudslay, part 2, 

plate xlviii 222 

LXXVIII. Inscription on the east side of Stela E. Quirigua. Maudslay, 

part 12, plate xxxn 232 

LXXIX. Plate lxii of the Dresden codex . 238 

LXXX. Plate lxix of the Dresden codex 240 

LXXXI. . Plate lxi of the Dresden codex 244 

LXXXII. Plate xxiv of the Dresden codex 284 

Figure 123. Part of inscription on "Monolithic Animal G", Quirigua. 

Maudslay, part 12, plate xltv 211 

124. Face numerals for 1 212 

125. Face numeral for 2 • 212 

126. Face numerals for 3 212 

127. Face numerals for 4 212 

128. Face numerals for 5 212 

129. Face numerals for 6 - 212 

130. Face numerals for 7 212 

131 . Face numerals for 8 212 

132. Face numerals for 9 213 

133. Face numerals for 10 213 

134. Face numeral for 11 213 

135. Face numerals for 12 213 

136. Face numerals for 13 ... 213 

137. Face numeral for 14 214 

138. Face numerals for 15 214 

139. Face numerals for 16 214 

140. Face numerals for 1 7 214 

141. Face numerals for 18 215 

142. Face numerals for 1 9 215 

143. Face numerals for 20 215 




Figure 144. Symbols for 0, or full count 216 

145. Symbols for full count, or 20 217 

146. Part of inscription on the east side of Stela C, Quirigua. 

Maudslay , part 11 , plate xix 217 

147. Part of inscription on the west side of Stela C, Quirigua. 

Maudslay, part 11 . plate xix 218 

148. Symbol for 13 cycles. Maudslay, part 10, plate lxxv, glyph 

C5 221 

149. Type of face numeral 222 

150. Part of inscription on the east side of Stela P, Copan 224 

151. Part of inscription on the east side of Stela P, Copan. 

Maudslay, part 4, plate lxxxix 225 

152. Glyph 33, west side of Stela F , Quirigua. Maudslay, part 12, 

platexL 227 

153. Part of inscription on the east side of Stela F, Quirigua. 

Maudslay, part 12, plate XL 227 

154. Part of inscription on the west side of Stela E, Quirigua. 

Maudslay, part 12, plate xxxi _ . _ 229 

155. Part of inscription on the west side of Stela E, Quirigua. 

Maudslay, part 12, plate xxxi _ . 230 

156. Part of inscription on the west side of Stela E, Quirigua. 

Maudslay, part 12, plate xxxi ' . 231 

157. Inscription at Xcalumkin, Yucatan. From a photograph by 

Mahler 253 

158. Part of inscription on the west side of Stela N, Copan. 

Maudslay, part 4, plate lxxix 254 

159. Lower division of plate lix, Dresden codex 259 

160. Great cycle symbols. Goodman, page 83 264 

161. Comb-like symbols for 20. Goodman, page 83 264 

162. Types of great cycle symbols from the inscriptions 266 

163. Types of the ahau (360) symbol 267 

164. Types of the katun symbol 268 

165. Types of the cycle symbol 268 

166. Column from plate xliii, Dresden codex 297 

167. Centerpieces of great cycle symbols 298 

168. Two symbols from a Chichen-Itza inscription 300 

169. Inscription on Stela 6, Copan. From photograph by Saville. 301 

170. Inscription on the nephrite stone in the Leyden Museum. .. 302 


By Cyrus Thomas 


When the paper entitled Mayan Calendar Systems, published in 
the Nineteenth Annual Report of the Bureau of American Ethnol- 
ogy, was written, the parts of Maudslay's work a relating to the ruins 
at Quirigua had not been received, and hence these important ruins 
could not then be considered, except so far as they were referred to 
by Goodman. b As these parts of Maudslay's work are now at hand, 
it is my purpose to supplement my previous paper by some notes on 
the inscriptions at Quirigua, and to discuss points omitted or but 
lightly touched in it. One of the points but briefly noticed is the 
value of the different face numerals. As was stated, the determina- 
tion of the value of these symbols necessitated a careful comparison 
of the series of the various inscriptions in which they are used, 
especially the initial series. This examination has been made, and 
the results are now given. 


The first inscription to which attention is called is that on the west 
side of Stela F. This is shown in the photograph (plate xxxix) and 
the drawing (plate XL) in part 12 of Maudslay's Archaeology, volume 
2, and in our plate LXXI. In regard to it Mr Goodman remarks as 
follows : 

Initial date: 54-9-14-13-4x17 — 12 Caban-5 Kayab. The period numbers here 
are expressed by face numerals. Following this date are fifteen indeterminable 
glyphs. They do not include the usual initial directive series, but they probably 
serve the same or a similar purpose, for we can distinguish a number of period 
symbols with accompanying numerals, though unable to determine their meaning 
here. Then comes a reckoning which reads, reversing the order of periods for 
convenience, as I shall do in all cases when necessary: 13-9x9, from 12 Caban-5 
Kayab, the initial date, to 6 Cimi-4 Tzec. 

The first, or initial, time series, 54-9-14-13-4-17, 12Caban 5 Kayab, 
is, as lias been explained in my preceding paper, to be interpreted as 

a Maudslay, A. P. Biologia Centrali-Americana: Archeology. London, 1889-1902. 
^Goodman, J. T. Archaic Maya inscriptions (appendix to the preceding). London, L897. 


204 MAYAN CALENBAK SYSTEMS [eth. ann.22 

follows: Fifty-fourth great cycle, 9 cycles, 14 katuns, 13 ahaus, 4 
chuens, and 17 days, to 12 Caban 5 Kayab, counting forward from 4 
Aliau 8 Cumhu, the first day of the fifty-fourth great cycle, as Good- 
man has numbered these supposed time periods. 

It is proper, however, to mention at the outset that the terms "great 
cycle," "cycle," "katun," "ahau," and "chuen" are used merely for 
convenience in comparisons with Goodman's renderings, and that I do 
not accept them as appropriate, or in any way adopt his theory that 
the} 7 denote real time periods, because I believe them to be nothing 
more than the orders of units in Mayan numeration; nor must it be 
understood that I accept his theory of a separate Mayan chronologic 
system. As the application of these terms has been fully explained 
in my previous paper, it is only necessary to restate here their numer- 
ical value : 

1 chuen 20 days (1X20) 

1 ahau 360 days (18X20) 

1 katun 7, 200 days (18X20X20) 

1 cycle. '. 144,000 days (18X20X20X20) 

The great cycle as given by Goodman equals 1,872,000 days or 18 x 
20x20x20x13, but should, as I shall endeavor to show, be counted 
as equal to 2,888,000 days, or 18x20x20x20x20. The number 54 
standing in the great-cycle place in the above series (54-9-14-13-4-17) 
is to be considered as having no numerical value ; it is not to be read 
"54 great cycles," but "the fifty-fourth great cycle" (according 
to Goodman's method of numbering these supposed time periods), 
while the other numerals, 9, 14, etc., are to be used as true numbers — 
that is, 9 cycles, 14 katuns, 13 ahaus, 4 chuens, 17 days — the 54 being 
entirely omitted from the calculation. The sum of the series will 
therefore be as follows, the day being the unit: 

9cycles (each 144, 000) 1,296,000 days (9X20X20X20X18) 

14 katuns (each 7, 200) 100,800 days (14X20X20X18) 

13 ahaus (each 360) 4, 680 days ( 13 X 20 X 18) 

4 chuens (each 20) 80 days (4x20) 

17 days 17 days 

i k 

Sum of the series 1 , 401 , 577 days 

After the initial series the next number-series (reversed), 13-9-9, or 
13 ahaus, 9 chuens, and 9 days, is found in the compound glyph num- 
bered 16 in Maudslay's drawing, the numbering of which has been 
retained in our plate lxxi. The date which follows — 6 Cimi 4 Tzec — 
is found in the right-hand portion of glyph 18 and the left-hand 
portion of glyph 19. 

As all the numbers of the initial series, including that attached to 
the month and day forming the terminal date, are face characters, 
and are considerably worn and dim, the question arises, How did 
Goodman ascertain their number value? 

Although some of these characters are so dim and imperfect that 



^ U 

> Ms) 

gtt^ \ I IM6&6Z rr— SV] 

^- — (rn — —j 












- ^ ■ ■ i ^ ' 



'3 ^^ 












j )L^^.Cj 











their details can not be traced with certainty, I will overlook this for 
the present and will try to get the data necessary to determine their 

Let ns suppose at first that the number value of no one of them has 
been ascertained. The first step will be to count back from the date 
following the next numeral series, in which the numbers are of the 
ordinary type. Although the symbol interpreted Tzec is too much 
worn to be determined from the photograph, I accept the drawing, 
which seems to indicate this month, as the artist had an opportunity 
of inspecting the cast. The date will therefore be 6 Cimi 4 Tzec. The 
preceding number series is 13-9-9, or 13 ahaus, 9 chuens, 9 days, and 
equals 4,869 days. By counting back from 6 Cimi 4 Tzec (year 1 
Akbal) we reach 12 Caban 5 Kayab (year 13 Ben), the concluding 
date of the initial series as given by Goodman. This, if the month 
symbol of the second date has been correctly interpreted Tzec, gives us 
the value of the number symbols attached to the first date, 12 Caban 
5 Kayab (glyphs 6 and 7, plate lxxi). Although these glyphs, as 
seen in the photograph, are scarcely distinct enough to be used in 
comparison, they are more clearly shown in the drawing, and present 
some characteristics which will assist us, especially that one (glyph 
7) denoting 5, attached to the month symbol, where the superfix is a 
form of the ordinary ahau symbol. As neither of these is repeated 
in the initial series, they afford us no aid in determining other face 
numerals of the series. 

It may be well, before proceeding farther with our examination of 
the series, to ascertain what data are necessary to determine the num- 
bers of the time periods in an initial series, and this can best be done 
by examples. Here we have, supposedly, as the initial date, 4 Ahau 
8 Cumhu (year 8 Ben), the first day of Goodman's fifty-fourth great 
cycle; and 12 Caban 5 Kayab is the concluding date of the series. 
That these two items are not sufficient to determine the intermediate 
time periods will be admitted without question. 

Let us suppose, as a means of further test, that the numbers of 
chuens and days, "4 chuens 17 days," given by Goodman, are correct. 
That 9 cycles, 14 katuns, 13 ahaus, 4 chuens, and 17 days, when counted 
forward from 4 Ahau 8 Cumhu, will bring us to 12 Caban 5 Kayab, 
as is maintained by Goodman, is true, as may easily be seen by mak- 
ing the calculation. 


9 cycles... 1,296,000 

14 katuns 100, 800 

1 3 ahaus 4, 680 

4 chuens 80 

17days... 17 

Total 1,401,577 

Subtract 73 calendar rounds 1 . 385, 540 

Remainder 16, 037 

206 MAYAN CALENDAR SYSTEMS [eth. ann. 22 

Subtracting from this remainder the 17 days which remain in the 
year 8 Ben, after 4 Ahau 8 Cumhu, and dividing the remainder by 
365, we obtain 43 years 16 months and 5 days. Counting forward 
this length of time (in the manner explained in my previous paper) 
from 4 Ahau 8 Cumhu, year 8 Ben, brings us to 12 Caban 5 Kayab, 
year 13 Ben. a 

The "calendar round ' : is, as has been explained in my previous 
paper, the term Goodman applies to the 52-year cycle, at the end of 
which period, counting from any point, the same date as that from 
which we count returns. The casting out of these calendar rounds, 
each of which amounts to 18,980 days, does not affect the result, as 
counting the remainder from the initial to the terminal date will give 
precisely the same result as counting the entire sum of the series — 
except that to determine the lapse of time, the number of }^ears 
covered by the calendar rounds cast out must be added. For example, 
in case of the above-mentioned series, as 73 calendar rounds were cast 
out, 73 x 52 years must be added to the result obtained by dividing 
the remainder by 365, in order to ascertain the real lapse of time from 
the initial to the terminal date. 

Having the date 12 Caban 5 Kayab and (supposed) the 4 chuens 
(or months) and 17 days, we turn to my condensed calendar or to 
Goodman's "Archaic Annual Calendar," and search through the 
tables of years until we find the year in which 12 Caban is the 5th 
day of the month Kayab. This in Goodman's tables is found to be 
the 51st year, or, in my table, the year 13 Ben. Counting back on 
the table of this year 4 months and 17 days, we reach 6 Ahau, the 8th 
day of the month Ceh, which, according to Goodman's scheme, will be 
the first day of an ahau. Turning now to Goodman's "Archaic 
Chronological Calendar" and to his 54th great cycle, we hunt for the 
place where 6 Ahau is the 8th day of the month Ceh. We find this 
in the 9th cycle, 14th katun, and looking at the column at the left 
margin we ascertain that it is the 13th ahau, which agrees exactly 
with the initial series as given above (54-9-14-13-4-17). 

This seems to be confirmatory; however, before accepting it as con- 
clusive let us examine a little further. Without any change, or sup- 
posed change, from the date and numbers of chuens and days used 
in the preceding calculation, we look farther in Goodman's "Archaic 
Chronological Calendar " to see if 6 Ahau 8 Ceh can be found else- 
where, confining our examination to his 54th great cycle. We do find 

it in the 13th cycle, 4th katun, 17th ahau, which gives the series 

Remembering that the 13th cycle, according to his scheme, is the 
first cycle of his great cycle, and must, therefore, be omitted from the 
calculation, and counting forward 4 katuns, 17 ahaus, 4 chuens, and 
17 days from 4 Ahau 8 Cuinhu, the first day of the great cycle, we 

a For condensed calendar and table of years see the end of this paper. 


reach 12 Caban 5 Kayab, the required date, as with the preceding 
series. Looking* farther we find 6 Ahau 8 Ceh in the 2d cycle, 12th 
katun, 6th ahau, giving the series 54-2-12-6-4-17, which also carries 
us to the proper date (12 Caban 5 Kayab). The date 6 Ahau 8 Ceh 
is also found in the 4th cycle, 19th katun, 15th ahau, and other places 
in the 54th great cycle, each of which gives the proper result. But 
this is not all, as we also find 6 Ahau 8 Ceh in the 53d great cycle in 
the 1st cycle, 7th katun, and 12th ahau, giving the series 53-1-7-12- 
4-17, which, counted from 4 Ahau 8 Zotz, the first day of the 53d 
great cycle, brings us to 12 Caban 5 Kayab, the required date. Other 
series which will give the proper result might be noted, but these will 
suffice to show that the initial and terminal dates and the chuens and 
days do not afford sufficient data by which to determine the series. 
It is necessary, therefore, to know the numbers attached to one or 
more of the other time periods of the series, and these must be ascer- 
tained in every instance by inspection and by a previously obtained 
knowledge of the value of one or more of the face numerals. 

Referring again to the initial series under consideration — 54-9-14- 
13-4-17, 12 Caban 5 Kayab — and holding to our assumption that the 
number of the chuens and days is correct, the date being satisfac- 
torily determined, we proceed to learn what additional data are neces- 
sarv to determine the series. 

If inspection and a knowledge of the face numbers prove the one 
attached to the cycle in this instance to be 9, then the series as given 
above is the only one that will agree with the data, and we are thus 
enabled to determine the value of the face numerals attached to the 
katun and ahau symbols; and should that giving the number of days 
be imperfect or obliterated, it would still be possible to determine the 
series, as the date with the other items mentioned (number of chuens 
and cycles) is always sufficient to fix the other numbers in the series. 
If the number attached to the chuens be unknown, then the series 
could not be determined with the other data mentioned. 

Suppose the number (9 in this case) attached to the cycle symbol 
to be imperfect or unknown, but that attached to the katun (14 in 
this instance) to be known, the series given above is the only one that 
will agree with the data. If the numbers attached to the cycle and 
katun be indeterminable, but that attached to the ahau symbol (13 in 
this instance) be known, the series can still be determined, and will 
be as given. It is apparent, therefore, that, with the initial and ter- 
minal dates and chuens and days known, the number attached to one 
of the other elements in the series is necessary in order to determine 
the series. It is also demonstrable that with these data the series 
can be at once determined by Goodman's tables, though this, as T 
shall show, does not prove that his theory of the Mayan time system 
or his method of numbering the cycles or great cycles is correct. 


Continuing our investigation of the data necessary to determine the 
series, still referring to the one under consideration, we will next sup- 
pose that the number of chuens can not be determined by inspection. 

The terminal date being given — 12 Caban 5 Kayab (which falls in 
a Ben year) — it is readily seen, by reference to Goodman's "Archaic 
Annual Calendar," 51st year, or to my condensed calendar, that it 
requires 17 days, counting back, to reach an Ahau which falls on the 
8th day of the month (Goodman begins the count with 20 Eb, but this 
gives Ben as the 1st day of the month, and the result is the same), 
hence the Ahau to be used depends on the number of chuens — if 
chuens 17 days, it will be — as seen by the table referred to — 8 Ahau 
8 Pax; if 1 chuen 17 days, then 1 Ahau 8 Muan; if 2-17, then 7 Ahau 
8 Kankin; if 3-17, then 13 Ahau 8 Mac; if 4-17, then 6 Ahau 8 Ceh; 
if 5-17, then 12 Ahau 8 Zac; if 6-17, then 5 Ahau 8 Yax; if 7-17, then 
11 Ahau 8 Chen; if 8-17, then 4 Ahau 8 Mol; if 9-17, then 10 Ahau 8 
Yaxkin; if 10-17, then 3 Ahau 8 Xul; if 11-17, then 9 Ahau 8 Tzec; 
if 12-17, then 2 Ahau 8 Zotz; if 13-17, then 8 Ahau 8 Zip; if 14-17, 
then 1 Ahau 8 Uo; if 15-17, then 7 Ahau 8 Pop; if 16-17, then 9 Ahau 
8 Cumhu; if 17-17, then 2 Ahau 8 Kayab. The fact that Ahau is 
the 8th day of the month in each case greatly limits the range of 

Suppose that, in addition to the terminal date, the numbers of 
cycles and katuns are also known (9 and 14 in this instance) ; the series 
can be definitely determined, and will be as given above. If the 
numbers of cycles (9) and ahaus (13) are known and the number of 
katuns is unknown, the series "54-9-14-13-4-17" will give the correct 
date, but there is one other — 53-9-13-13-13-17 — which will also give 
the correct date, 12 Caban 5 Kayab. In this case the correct deter- 
mination of the series depends on the initial day of the great cycle, to 
which attention will be called farther on. 

We next take the case where, in addition to the dates and the 
number of days, the numbers of katuns and ahaus are known, and 
the number of cvcles is unknown. In the series under consideration 
the number of katuns is 14, of ahaus 13. These data are sufficient to 
determine the series, and in this instance the result is as given above. 

The next inquiry relates to the data necessary to determine the ter- 
minal date where this can not be recognized by inspection, or where 
that given is erroneous. Where neither the day nor the day of the 
month is known, it is necessary to have the entire numeral series — 
that is, 54-9-14-13-4-17, in the example we have been using — in order 
to determine the date. If the day of the terminal date of the series 
can be ascertained by inspection, then the date can be determined 
without knowing the number of days; thus 54-9-14-13-4-?, ? Caban 
? (month) will be sufficient to ascertain that this terminal date is 12 
Caban 5 Kayab. Turning to Goodman's "Archaic Chronological 
Calendar," 54th great cycle, 9th cycle, 14th katun, 13th ahau, we find 







6 Ahau 8 Cell. Searching through his "Archaic Annual Calendar'' 
we find that Ahau 8 Ceh occurs only in the 51st year, and that Caban 
is the 5th day of the month in this year. Counting forward 4 months 
from Ceh brings us to Kayab, where 12 Caban is the 5th day. We 
thus ascertain that 12 Caban 5 Kayab is the date sought. 

If the number of days, the name of the day of the terminal date, 
the month, and da}" of the month be unknown— thus in our example 
54—9-14—13-4-?, 12 (day) ? (month) — it is possible to limit the result to 
one of two days, in this case to 12 Kan 12 Pax, or 12 Caban 5 Kayab. 
In the first case, the number of daj's will be 1, and in the second 
17. If the number of chuens and the day and month of the date 
be unknown, but the number of the day and the day of the month 
known, the date can be determined. 

There are occasional side aids which may be taken advantage of in 
the investigation of the face numerals. One example which we will 
notice, bearing on the series which has been under consideration 
(initial series 54-0-14-13-4-17, west side Stela F, Quirigua), is as 
follows: The initial series on the west side of Stela E, Quirigua (plate 
lxxii), is, ordinary numerals being used throughout, and all distinct, 
54-0-14-12-4-17, 12 Caban 5 Cayab. This is identical with the other 
series, except that there are only 12 ahaus, while in the other there 
are 13. 

Although all that is positively known in regard to the first series 
(so far as our present investigation has extended) is the initial and 
terminal dates, the number of the days, and the day of the month on 
which the Ahau falls, we also know that the series as given above 
will agree with these items. If the 12 ahaus in the second series 
given above should, in fact, be 13, there will be perfect agreement 
with that on the west side of Stela F. It is evident from what has 
been shown above that, with all the items of the series save one being 
known, that item can be determined although wholly obliterated or 
incorrect. Enough is given to show that, counting back 4 months 
and 17 days from 12 Caban 5 Kayab, Ave reach (3 Ahau 8 Cell. By 
calculation, or by referring to Goodman's "Archaic Chronological 
Calendar," 54th great cycle, 0th cycle, and 14th katun, it is seen that 
6 Ahau 8 Ceh can onlv be in the 13th ahau, and is not found in the 
same cycle and katun in either the 53rd or 55th great cycle. The 
question as to whether Goodman's tables cover the range of the initial 
series will be considered farther on, when we have investigated more 
series. However, the fact that the series on the west side of Stela E, 
when the number of ahaus has thus been corrected (as calculation 
also shows 12 to be wrong), agrees p: ecisely with the rendering given 
of that on the west side of Stela F is not proof that this rendering is 
correct, it only adds a degree of* probability, supposing that Goodman 
has based his determination on an examination of the face characters. 
The fact maybe noted, also, that some two or three other inscriptions 

22 eth— 04 14 

210 MAYAN CALENDAR SYSTEMS * [eth. ann. 22 

at this place, where the numerals are of the ordinary form, commence 
with 9 cycles. 

As the numbers in the inscription on the east side of Stela E are 
all of the ordinary form, nothing in regard to the face numerals can 
be learned from it. 

The numbers in the initial series on the east side of Stela F (plate 
lxxiii) are all face characters, including those attached to the termi- 
nal date. Goodman interprets them as follows: "54-0-16-10-18-20, 
1 Ahau 3 Zip. " As will be seen by reference to my former paper, the 
18 chuens 20 days are to be understood and counted as chuens 
days, and we shall hereafter write them so. Goodman omits, prob- 
ably by printer's mistake, the 9 cycles, but the other numbers which 
he gives make them necessary. 

As none of the numbers in this case correspond with any on the 
west side of the same Stela, excepting the 9 cycles, the glyph for 
which is too nearly obliterated for determination, we gain nothing by 
comparison; and nothing can be learned from other inscriptions of 
this locality which present no face numerals; these are passed over 
without notice. 

Turning to plate xliv in part 12 of Maudslay*s work we find draw- 
ings of the inscriptions on the "Monolithic Animal G." As the 
numerals in the initial series are face characters, with the exception 
of that attached to the month of the terminal date, and have not 
been noticed by Goodman in his work, I call attention to them 
(figure 123). As the cycle in most of the initial series at Quirigua 
appears to be numbered 9, we will assume that to be the number in 
this case. But this is not a mere assumption without any other basis, 
as the .glyph is not inconsistent with that on the west side of Stela F 
and agrees with the type given (see figure 132) in having the circle of 
dots on the cheek. Although this does not amount to demonstration, 
it renders the interpretation highly probable. Having determined 
the cycle our examination is very much restricted. However, as we 
know as yet no way of determining the great cycle by an inspection 
of the* symbol, our examination must extend to the three given by 
Goodman. But without other data the examination on this line is 
vain. Examining the series, we notice that the face glyph attached to 
the katun symbol immediately under the cycle is partially obliterated 
and as yet is unknown. Passing to the upper pair in the next group 
to the right hand, we notice that the numeral resembles somewhat 
closely that attached to the month (glyph 7) of the terminal date in 
the inscription on the west side of Stela F (plate lxxi), which was 
found to denote 5. The symbol on the monolith differs in having the 
skeleton jaw, which Goodman says denotes 10, though we have not as 
yet found the proof of this, and we therefore assume that it denotes 
15 (10+5) (see figure 1385). The hand on the face immediate^ 
below, which is attached to the chuen glyph, as also on the glyph 








attached to the symbol for days in the upper pair of the group to the 
right, denotes, according to Goodman, full count or 20 when days and 
18 when chuens (see figure 143). However, 
I consider it, as heretofore stated, a symbol 
for naught. Immediately below the latter 
is the day (probably Ahau) of the terminal 
date, with the face symbol for 5, already 
determined, prefixed (figure 128 a). The 
first glyph of the lower pair of the group to 

the ri^ht has the ordinary character for 3 3 

prefixed. This we take to be the month ;_, 
symbol, though it is unusual and indeter- 

minable by inspection. The series, there- & 

fore, so far as made out, is as follows: o 

54?_9_?_1 5_0-0, 5 Ahau 3 (month). B 

It is evident that the 5 Ahau of this series §. 

must be the beginning day of an ahau, as § 

there are neither chuens or days, and hence g 

it should be found in Goodman's "Archaic = 

Chronological Calendar." Turning to this g 

publication, we find that 5 Ahau 3 (month) £• 

can occur as the beginning of the 15th ahau a 

in the 9th cycle in the following places only — & 

53d great cycle, 9th cycle, 17th katun; 54th | 

great cycle, 9th cycle, 4th and 17th katuns. p 

In the first it falls in the month Pop, in the ,© 

second in Yaxkin, and in the third in Muan. g; 


As the month symbol, so far as it remains, g 

does not admit of interpretation as the first g 

or second of these, we conclude that it must g 

stand for the third if the date is within the £ 

range of Goodman's calendar. This gives as ^ 

the series 54-9-17-15-0-0, 5 Ahau 3 Muan, 1 

which works out correctly by calculation. » 

The "full count" or "naught" symbols %■ 

require some discrimination in our attempts £ 

to interpret them. In a series given by § 
Goodman, as 54-9-17-15-18-20, or 54-13- 
20-20-18-20, 18 and 20, being so-called "full 
counts," should in every instance be counted 
as naught, and the cipher (0) should be in- 
troduced in their place; and this is true in 
every case where the symbols are used to 
represent prefixed numbers, except in one 
place. Where they are used to denote the day of the month, as 5 Kb 
20 Zotz, they denote 20, but there are special characters used for this 



[ETH. ANN. 22 

purpose, as is shown in figure 145. It appears probable also that the 
hand across the jaw in the face-forms of the cycle and great cycle is 
to be interpreted as indicating the use of 20 as a multiple, though in 
face-forms of prefixed numbers it undoubtedly signifies naught. 
Goodman is possibly right in insisting that these are not absolutely 
naught symbols, as is our 0, but are used to indicate that the count in 
the given denomination is complete and has been carried into the next 
higher denomination. Nevertheless they are — with the exceptions 
mentioned — equivalent to naught and must be so considered and used 
in calculating time and numeral series. 

be d 

Fig. 124. Face numerals for 1. 

Fig. 125. Face nu- 
meral for 2. 



Fig. 126. Face numerals for 3. 

Fig. 127. 

b c 

Face numerals for 4. 

s a 

b c d 

Fig. 128. Face numerals for 5. 

a b 

Fig. 129. Face numerals 

for 6. 

a bed 

Fig. 130. Face numerals for 7. 


Fig. 131. 

b c 

Face numerals for 8. 

I insert here, in figures 124 to 145 inclusive, the types of face 
numerals selected by Goodman from the inscriptions. I have found 
them to be correct, with some two or three exceptions in regard to 
which there is considerable doubt; these will be noticed in the proper 
connection. Some additional examples will appear as we proceed. 

The next inscription of this locality to which attention is called is 
that on Stela J (see Maudslay's drawing, part 12, plate xlvi, our plate 
lxxiv). All the numbers of the initial series except that of the day 
of the month in the terminal date are face characters. The series 




as given by Goodman is as follows: ?-9-l 6-5-0-0, 8 Ahau 8 Zotz. 
The number of the great cycle is omitted, though it is necessarily 54 
according to his system. He says there are no other reckonings in 
the inscription, but this is a mistake, as there are two more numeral 
series, each followed by a distinct date. These, however, afford no 
assistance in determining the initial day, as they do not connect with 
it; moreover, a large number of glyphs intervene. 

All the evidence bearing on the value of the face numerals in this 
instance may be stated as follows: the symbol connected with the 
cycle, interpreted 0, shows the distinguishing features of the others 


6 c 

Fig. 132. Face numerals for 9. 

c d 

Fig. 133. Face numerals for 10. 


Fig. 134. Face nu- 
meral for 11. 


b c d 

Fig. 135. Face numerals for 12. 


b c d 

Fig. 136. Face numerals for 13. 

noticed which are interpreted 9. This, taken in connection with the 
fact that most of the inscriptions of this locality begin the initial 
series with 0, renders the interpretation probable. We have as yet 
no evidence that 10 is the correct rendering of the character attached 
to the katun glyph, Goodman's example (figure L39&) being the very 
symbol found here; but the 5 attached to the ahau glyph agrees with 
that determined from the inscription on the west side of Stela F, and 
therefore may be accepted as correct. Hie face number attached 
to the day (Ahau) of the terminal Hate, which is interpreted 8, is as 



[ETH. ANN. 22 

Fig. 137. Face nu- 
meral for 14.. 

Fig. 138. Face numerals 
for 15. 

yet undetermined in our investigation; it is c of our figure 131. The 
day of the month and the month (8 Zotz) are distinct and easily recog- 
nized, the number being of the usual form. In regard to the chuen 

and day symbols, all we can say is that the hand across 
the face which appears to indicate full count (18 and 
20) or naught (0) is seen in each of the attached 
glyphs. Assuming this to be correct, it follows that 
the date 8 Ahau 8 Zotz must be the first day of a 5th 

Turning in Goodman's Archaic Chronological Cal- 
endar to the 9th cycle of the 53d great cycle, we find that 8 Ahau 8 
Zotz is not the beginning of any 5th Ahau in this cycle nor in the 9th 
cycle of the 55th great cycle, but is the begin- 
ning of the 5th ahau of the 16th katun in the 
9th cycle of the 54th great cycle. Even omit- 
ting the number of the day Ahau we can reach 
the same result from the data given, and that 
result only. The evidence therefore appears 
to be sufficient. This gives one example of the 
face character for 16 (see figure 139 b). As to 
the value and reliability of Goodman's tables 
in the respect noticed I will speak hereafter ; at present I as- 
sume them to be reliable, and I may state here that they may be 

accepted, so far as our present 
tests are concerned, as correct 
in regard to the relation of 
the several time periods up 
to and including the cycle — 
without, however, accepting 
his theory in regard to the 
great cycle or the number of 
cycles forming one of these great periods. 

We must therefore accept as determined with reasonable certainty 
the value of the following face 
numerals: that on Stela J 
(glyph 1, Maudslay's plate 
xlvi, our plate lxxiv) pre- 
fixed to the cycle glyph, in- 
terpreted 9; that (glyph 3) 
affixed to the katun glyph, 
interpreted 16; that (glyph 5) 
prefixed to the ahau glyph, 

interpreted 5; those (glyphs 7 and 9) prefixed to the chuen and day 
glyphs, interpreted full count or naught; and that (glyph 11) prefixed 
to the day of the terminal date (Ahau, in this instance), interpreted 8. 
One distinguishing characteristic of the symbol for 9 is the circle of 

Fig. 1 39. Face numerals for 16. 


Fig. 140. Face numerals for 17 




S^SfeW ■ 

' k^ ^^ > 

Ct ■' *> ' " \ •> > * 

< <4 I / 


7 V^£-~^>> i ^ 


£T? ^ 1 ' (/f\ , ^ 



dunlin i^.'—' -^ V^^Avfc_ - ^A^___ < »^' 
ft N \ f \ / \ 

f A 3 j w^^V 

Q ^M\ q 

v €~r 








dots on the cheek (figure 132) ; two characteristics of the symbol for 
16 are the skeleton jaw and the hatchet in the eye (figure 139) ; those 
of the symbol for 5 are the ahau symbol on the head and the absence 
of the skeleton jaw (figure 128); that of the symbol for full count or 
naught is the hand across the face or lower jaw (figure 143) ; those of 


b c 

Fig. 141. Face numerals for 18. 


the symbol for 8 do not appear to be well defined — Goodman say? they 
are the lobed ear ornament projecting on the cheek and the foim of 
the forehead ornament, but neither of these appears to be exceptional. 

It should be stated that by counting forward in each of the given 
examples from the initial date (4 Ahau 8 Cumhu) the number of days 
indicated by the numeral series we will 
reach the terminal date. 

Our next reference is to the inscription on 
Stela A, Quirigua, the drawing of which is 
given in plate vn, part 11 of volume II, 
Maudslay's Archaeolog} 7 (our plate lxxv). 

In this instance the numerals attached to 
the cycle, katun, and ahau, and the month 
of the terminal date of the initial series are 

of the ordinary form, and those attached to the chuen, day and the 
day of the terminal date are either face forms or unusual forms. The 
series as given by Goodman is 54-9-17-5-0-0, 6 Ahau 13 Kayab, 
which is certainty correct, as the data given are sufficient, as has 
been shown, to determine the series. It agrees with Goodman's tables 
and also with calculation. 

Fig. 142. Face numerals for 19. 


b c <l 

Fig. 143. Face numerals for 20. 

By this we ascertain that the unusual numerals (glyphs 4 and 5) 
prefixed to the chuen and day symbols, each of which consists of a 
scroll above, a hand in the middle, and a bean-shaped character below, 
denote naught (figure 144, number 7). The face numeral prefixed to 
the day Ahau (figure 129 6) resembles that denoting 1G (see figure 



[ETH. ANN. 22 

139 b, c), excepting that it is without the skeleton jaw, thus appar- 
ently confirming Goodman's statement that this characteristic has 
the value of 10. In figure 144 are shown some forms of the symbols 
for naught (0). Numbers 1, 2, 3, 4, 5, and 6 in some of the types are of 
frequent occurrence in the inscriptions, as are also numbers 7 and 8. 



<^> ^^<£B>-^^><^E>> <p£> <z^> <$£> 


Fig. 144. Symbols for 0, or full count. 

Numbers 9 and 10, which show the hand across the lower jaw, also 
represent a common type. Number 12 has been found only in the 
inscription on the Palace steps, Palenque. Number 11 is from Mono- 
lithic Animal B, Quirigua, and numbers 13 and 14 are from Stela D, 



r (®J?ii 


y /? . J ~— 










IJcJ^A 1 






Fig. 145. Symbols for full count, or 20. 

Copan. The small figures of number 15 are from the Dresden codex, 

and represent a common type; the slight variations in detail are 

numerous and appear to 

have no significance. 
In figure 145 are shown 

the symbols for full count, 

or 20, not shown in figures 

143 or 114. A, b, c, and d 

are more or less common 

in all the codices; e is 

from the Dresden codex; 

/, g, h, and i are from the 

left slab, Tablet of the 

Cross, Palenque. 

The inscription on the east side of Stela C presents some particu- 
lars worthy of notice (see figure 146). The prefix to the cycle sym- 
bol is 13 in the ordinary form; those to 
the katun, ahau, and day are of the 
coil and hand form, above described as 
indicating full count, or, in other words, 
naught ; that to the chuen is of the usual 
form for full count in the inscriptions 
(see number 3, figure 144). The date is 
4 Ahau 8 Cumhu with ordinary numer- 
als. Therefore the series, according 
to Goodman's method of writing, will 
be ?-13-20-20-18-20, 4 Ahau 8 Cumhu, 
which is as he gives it, excepting that 
he places it in his fifty-fourth great 
cycle. Our method of writing it would 
be 53-13-0-0-0-0, 4 Ahau 8 Cumhu. 
I give 53 as the great cycle, accord- 
ing to Goodman's method of number- 
ing these periods, as by counting back 
13 cycles, or 1,872,000 days, from 4 
Ahau 8 Cumhu we reach 4 Ahau 8 Zotz, 
the first day of his fifty-third great 
cycle. His remark in regard to it is: 
"This date is the beginning of the fifty- 
fourth great cycle." As he interprets 
the great cycle 54, he virtually makes 
the series 54-0-0-0-0-0. It must be 
borne in mind, as will be seen by ref- 
erence to my former paper, that instead 

of counting 20 cycles to the great cycle, following the vigesimal 

system, which I believe to be correct, he counts 13. However, this 

Fig. 146. Part of inscription on the east 
side of Stela C, Quirigua. Maudslay, 
part 11, plate xix. 



[ETH. ANN. 22 

subject will again be referred to. At any rate, we find further con- 
firmation of the signification of the number symbol — the combined 
coil, hand, and bean shaped character — in this inscription. 

The inscription on the west side of Stela C (figure 147 a) is inter- 
preted by Goodman as follows: 9-1-0-0-0, 6 Ahau 13 Yaxkin, the 
number of the great cycle being omitted. As the numerals attached 
to the cycle, katun, and day and month of the terminal date are of 

the usual form, and the symbols for 
full count, or naught, attached to the 
ahau and month glyphs are of the 
usual type, we have sufficient data to 
determine the face character attached 
to the day glyph. Omitting from 
consideration the number attached 
to the day symbol and counting back 
from 6 Ahau 13 Yaxkin, yearlLamat, 

a b 

Fig. 147. Part of inscription on the west side of Stela C, Quirigua. Maudslay, part 11, plate xix. 

9 cycles and 1 katun, or 1,303,200 days, according to the method given 
in my former paper, we reach 4 Ahau 8 Cumhu, the first day of the 
fifty- fourth great cycle. Turning to Goodman's Archaic Chronological 
Calendar, to the ninth cycle of the fifty-fourth great cycle, we find 
that 6 Ahau 13 Yaxkin is the first day of the first katun. Thus it is 
seen that both the tables and calculation agree with the interpreta- 
tion of the series. By this we have a further confirmation of the 


interpretation full count (or properly 0) of the face numeral with the 
hand over the lower jaw. It may be mentioned here that Maudslay 
agrees with me in designating these so-called "full counts" as given 
by Goodman as "no count," or, in other words, naught (see his text, 
part 11, page 9). 

Referring to inscriptions in other localities, the following facts are 
noted in reference to the value of the different face numerals : the initial 
series of the Foliated Cross at Palenque (see figure 2, previous paper) 
appears to be as follows: 54-1-18-5-4-0 to 1 Ahau 13 Mac. Follow- 
ing this date, after some intervening glyphs, is the brief numeral 
series 14 chuens 19 days, immediately after which comes the date 
1 Cauac 7 Yax. 

Counting back 14 chuens 19 days from the latter date, we reach 1 
Ahau 13 Mac, the terminal date of the initial series. This gives the 
value 1 to the face gtyph attached to the Ahau symbol. This face 
glyph (figure 124 b) agrees in its features, excepting the ear pendant, 
with the face glyph attached to the cycle symbol (figure 124 a), show- 
ing it to be 1, which agrees with the above interpretation. As the 
face glyph attached to the ahau period symbol agrees with the symbol 
we have heretofore interpreted 5 (see figure 128 a); and the number 
attached to the month symbol is of the ordinary form; and that 
attached to the day glyph has the hand across the lower jaw, we have 
the following numbers of the series: ?-l-?-5-?-0, 1 Ahau 13 Mac. 

These items are not sufficient to give the remaining numbers of the 
series; but assuming that it falls in the 54th great cycle, as is most 
probable, the other numbers will be as given above. As the face 
character attached to the chuen symbol, interpreted 4 (figure 127 6), 
presents some features of the one interpreted 4 on the west side of 
Stela F at Quirigua (left part of glyph 4, plate lxxi), and this will 
suffice to determine the other numbers, we are perhaps justified in 
concluding that the series is given correctly. That the face character 
attached to the katun symbol (figure 155 c), which is interpreted 18, is 
some number greater than 10 is shown by the skeleton jaw. 

Turning to the inscription of the Temple of the Sun, as shown in 
Maudslay's plate lxxxix, part 10 (see plate xli, Nineteenth Annual 
Report of the Bureau of American Ethnology, 1900), where the num- 
bers of the initial series are all face characters except those designat- 
ing the day of the month in the terminal date, we will try to determine 
them from the data so far obtained. As those attached to the cycle 
(figure 124 c), katun (figure 141c), and ahau (figure 128 6) symbols 
are evidently the same as those in the inscription of the Foliated 
Cross, and the day of the terminal date is ? Cimi 19 Ceh, we have the 
following items of the series: ?-l-18-5-?-?, ? Cimi 19 Ceh. 

These data are not sufficient to determine the remaining numbers. 
One other item is necessary for this purpose. Assuming the great 
cycle to be that commencing with the day 4 Ahau 8 Cumhu, the 

220 MAYAN CALENDAK SYSTEMS [eth. ann. 22 

so-called 54th, the remaining numbers may be determined thus: Cimi 
may be the 19th day of the month only in the years in which Ahau 
is the 13th day of the month. By turning to Goodman's "Archaic 
Chronological Calendar," 54th great cycle, cycle 1 and katun 18, we 
see that the 5th ahau begins with the day 12 Ahau 13 Chen. Turning 
to his "Archaic Annual Calendar," we find that 12 Ahau 13 Chen 
falls in the year he numbers 34 (equivalent to the year 9 Lamat in 
my condensed calendar). Cimi is the 19th day of the month in this 
year, but the month can not be determined until the day number 
attached to Cimi is ascertained. As the face numeral attached to 
the chuen symbol in the inscription is without the skeleton jaw we 
infer that it does not exceed 9, and as it has none of the signs of full 
count or naught it can not be 0. As Cimi comes 6 days after Ahau, 
then we must count forward in the table of the year 34 until we 
reach the 19th day of the month Ceh. This count we find to be 3 
months and 6 days, and the number attached to Cimi is 13. There- 
fore the entire series is 54-1-18-5-3-6, to 13 Cimi 19 Ceh, which is as 
it is given by Goodman. The weak point in this solution is the 
assumption of the 54th great C3 T cle. Even without this, we can, by 
a range of nine trials, determine that no other numbers than those 
given can be found within the scope of Goodman's three great cycles 
(53d, 54th, and 55th), but this, though strengthening the conclusion, 
is not absolute demonstration, as the objection to his method of 
counting the cycles, hereafter noticed, and the uncertainty as to the 
scope of his tables, come into the problem. As will be seen later, the 
only certainty in regard to the tables of his "Archaic Chronological 
Calendar" is the orderly and correct succession of dates and periods 
and the fact that 4 Ahau 8 Cumhu is the first day of a great cycle. 
Assuming for the present that the series has been correctly deter- 
mined, we gain evidence as to the value of two additional face numer- 
als, 3 (figure 126a) and 6 (figure 129a). 

Goodman's interpretation of the initial inscription of the Tablet of 
the Cross, which is 53-12-19-13-4-0, 8 Ahau 18 Tzec, is not satisfac- 
tory. The face numeral attached to the cycle symbol, which he 
interprets 12 (figure 135 a) has, as a superfix, a figure very much like 
the superfix to the face character which he has correctly interpreted 
5 (as is shown by the evidence I have presented) (figure 128 a). In 
his representation of face numerals no one save those denoting 5 or 
15 have a superfix of this kind, excepting one for 12, and that one 
is the character of this inscription (figure 135a). Moreover, it lacks 
the skeleton jaw, which is true of some others above 10 as given by 
him. As has been shown in my previous paper, where this inscrip- 
tion is discussed at length, and as is admitted by Goodman, there is 
no connection between the terminal date of the initial series and any 
of the dates which follow, if the numeral series which intervene be 
taken as given in the inscription. 




In addition to the suggestions offered by Goodman and those pre- 
sented in my previous paper in regard to correcting the manifest 
error somewhere in these series, the following is added as a possible 
solution : Change the terminal date of the initial series from 8 Aliau 
18 Tzec to 1 Ahau 8 Muan, and the following numeral series will then 
connect the succeeding dates with it, and the 1 Ahau 18 Zotz will come 
1-8-0 (1 ahau 8 chuens) or 520 days after the terminal date of the 
initial series, instead of being placed back of it as 
Goodman's correction requires. This, however, will 
slightly change the initial series from the numbers 
given by Goodman. By referring to the inscription 
as given in Maudslay's drawing, we notice at Co the 
symbol for 13 cycles (figure 148). As this is not 
connected with a series, and follows immediately 
after the date 4 Ahau 8 Cumhu, we are justified in 
interpreting it as an indication that up to this point 
13 cycles have been passed over from the initial date 
of the inscription, which must be 4 Ahau 8 Zotz. 
is correct. Subtracting the series 8-5-0 (ID 2C) from 13 cycles the 
remainder is 12-19-11-13-0. 


Fig. 148. Symbol for 
13 cycles. Mauds- 
lay, part 10, plate 
lxxv, glyph C 5. 

The calculation 


If this correction be justified the initial series will be 53-12-19-11- 
13-0, 1 Ahau 8 Muan, which will fit into Goodman's tables. The chief 
objection to this is that it compels us to assume that the aboriginal 
artist made a mistake in his calculation, as the month symbol is 
clearly Tzec and the face numeral shows the skeleton jaw, indicating 
that the number as given is above 10. However, we must admit that 
the error has not, as yet, been satisfactorily explained, and conse- 
quently the value of but two of the face numerals — those attached to 
the cycle and katun glyphs — can be determined by the inscription. 
Twelve (see figure 135a) for the cycle and 19 (figure 142a) for the 
katun, as given by Goodman, must apparently be accepted on any 
theory as to the correction. It will be observed th ;t the s} r mbol 
attached to the ahau glyph, which Goodman interprets 13 (figure 136 a 7 ), 
is widely different from any of the other symbols for 13 given by 
him, as is seen by reference to our figure 13G, which is a copy of the 
examples given by him on page 49 of his work. So far, therefore, 
as comparison shows, it may as well be interpreted 11 as 13; but, in 
fact, is more like 19 (see figure 142) than either. Nor can his inter- 
pretation (4) of the character attached to thechuen symbol be clearly 
sustained by comparison, though it must be conceded that it does not 
resemble; the determined types of L3. 

The initial series on stela I) of the Copan Inscriptions (Maudslay, 
plate XLVIII, part 2, our plates LXXV] and LXXVIl) is peculiar in 

222 MAYAN CALENDAR SYSTEMS [eth. ann\ 22 

having the usual face characters replaced by full forms. The cycle 
symbol (glyph 1) is composed of a human figure (the numeral) 
and a bird apparently of the parrot species (the cj^cle); the katun 
(glyph 2) of the human form (the numeral) and a bird of a rapacious 
species (the katun); the ahau (glyph 3) of the human form (the 
numeral) and a nondescript animal (the ahau) ; the chuen (glyph 4) of a 
human form (the numeral) and a frog-like animal (the chuen) ; the 
day (glyph 5) of two human forms, that to the right with the monkey- 
like face turned backward (the day) ; the day of the date (glyph G) 
(presumably Ahau) of a human form (the numeral) with a cartouch 
inclosing another form (the Ahau); the month of the date (glyph 13, 
plate lxxvii) of a human form (the numeral) and a full-formed leaf- 
nosed bat (the month). 

Goodman's interpretation of the series is as follows: 54-9-5-5-0-0, 
4 Ahau 13 Zotz. The dots on the chin of the human face of the 
cycle symbol (plate lxxvi, glyph 1) and other characteristics prob- 
ably justify us in interpreting it as 9. The hand across the lower 
. jaw in the chuen symbol (plate lxxvi, glyph 4) and 
day symbol (plate lxxvi, glyph 5) indicate full count 
or naught (0). But Goodman's rendering 5 and 5 of the 
number characters of the katun (plate lxxvi, glyph 2) 
and ahau (plate lxxvi, glyph 3) symbols is question- 
able, as the skeleton jaw denoting 10 is quite distinct 
in the former and is not present in the latter. The 
rendering is therefore inconsistent with Goodman's own 
statements in regard to the characteristics of the face 
Fig. 149. Type numerals, and must have been reached in some other 

way than by inspection of the glyphs. If the figures 
i with ahau symbols on the head are face numerals, and this must be 
admitted, then that of the katun (gl} r ph 2) should be 15, and that of 
the ahau (glyph 3) should be 5, if Maudslay's colored drawing is cor- 
rect. However, it must be admitted that the drawing of the face 
numeral prefixed to the katun symbol is very doubtful. In figure 
149 is given a drawing of the head alone, made from Maudslay's plate 
xliv, which is the autotype of the same inscription. 

This inscription is the most interesting one in some respects that 
has been found in Mayan ruins. Entire bodies, instead of conven- 
tional heads, are given, and though they are to some extent gro- 
tesque, yet they seem to indicate the aboriginal idea of the origin of 
these symbols. Maudslay's happy idea of distinguishing the prefixed 
numerals from the period symbols (cycle, katun, etc.) by difference 
in color brings out very clearly the forms and characteristics of the 
latter symbols. The cycle and katun symbols are both rapacious 
birds; the former owl-shaped, with a crest; the latter eagle-shaped, 
with feathers hanging over the front of the head. The ahau symbol 
is the skeleton form of a nondescript bird-like animal with a large 







•-) _d& ^ 



^ I OvC-v,^ 

,x '-;v 


■ ^r^ u 





..^r j 


^ v 

~-A~*S s 








■s— - 

^S3 i 


A J ^ 



'dm \ 

r'— - 


kf Af * 




J ; 



■ %%^/-~ 

■ ■ 

r ~ m 


- \ 




- — r-\ 

■- v . 





fang; the chuen glyph is a frog-like animal. The month symbol of 
the date (glyph 13, plate lxxvii) is, as stated above, a leaf-nosed bat 
with a human face. As the name of the latter, Zotz, or "Bat," cor- 
responds with the form, it is possible that the forms of the other sym- 
bols have some reference to the names. However, I am unable to 
point out this reference; though possibly as " uinal " in Maya sig- 
nifies "month" or "period of 20 days," and "no" "frog," the sym- 
bol may have some reference to the name. Be this as it may, it will 
be seen by reference to figures 163 and 164, showing the types of the 
ahau and katun symbols, that the face forms retain to a large extent 
the bird-like features, one of the katun symbols, figure 164a, having 
the feather fringe over the forehead. We notice also in some of the 
symbols of both the ahau and katun little patches of cross-hatching, 
which are feather marks in the full forms of Stela D. 

These facts are noticed in passing merely to call the attention of 
students to them as possibly forming some clew to the relation between 
these symbols and what is represented by them. 

Attention is called next to the inscription on Stela I, Copan. The 
numerals attached to the cycle, katun, ahau, and chuen symbols are 
of the ordinary form ; that to the day glyph is of the disk and hand type 
(figure 111) denoting naught (0); and that to the day (Ahau) of the 
terminal date, the face charact r with the ahau headpiece denoting 
5. Whether the month symbol is distinguishable, or is one of the 
obliterated glyphs which follow, as Goodman asserts, is doubtful. The 
series is therefore ?-9-l 2-3-1 4-0, 5 Ahau ? (month). Goodman sa} T s 
54-0-12-3-14-20, 5 Ahau ? (month), leaving the month blank, but 
adds that we know it must be 8 Uo. 

The correctness of the last statement may be questioned on the fol- 
lowing grounds : Taking, in Goodman's own tables, the 55th great cycle, 
9th cycle, 12th katun, and 3d ahau, we find that the first day of this 
ahau is 11 Ahau 8 Uo; by counting forward 14 months from this date 
we reach 5 Ahau 8 Pax, a result which calculation shows to be cor- 
rect, the initial date of this great cycle being 4 Ahau 3 Kankin. The 
positive determination depends therefore on the proper determination 
of the great cycle, or of its initial day, for his numbering of these 
supposed periods, as we shall soon see, is without proper grounds. 

The initial series of the inscription on the east side of Stela P (figure 
150), same locality as the preceding, is given as follows : 54-0-9-10-0-0, 
2 Ahau 13 Pop. The numbers attached to the cycle, katun, and ahau 
are face characters, those attached to the chuen and day symbols are 
of the type shown in numbers 4, 5, and 6, figure 144, but much abbre- 
viated, and those of the terminal date are of the ordinary form. The 
month symbol, which Goodman interprets Pop, is apparently a 
variation of the usual type. As enough of the prefix to the ehuen 
symbol remains to indicate full count or naught (0), it may be 
assumed that the prefix to the day symbol, of which there seems to 



[ETH. ANN. 22 

be a slight remnant, is the same; therefore the terminal date will be 
the first day of an ahau. The skeleton jaw in the prefix to the ahatt 
symbol, not well shown in Mandslay's drawing (plate lxxxix of his 

A b 

Fig. 150. Part of inscription on the east side of Stela P, Copan. 

work, part 4), but distinct in his photograph, would indicate 10 or some 
number above 10 (see figure 150). The face numerals of the cycle and 
katun are evidently the same, and one of them shows quite distinctly 




the circle of dots on the cheek, indicating 9 (see figure 132). There- 
fore the series so far as satisfactorily made out — assuming the number 
attached to the day Ahau to be 2 — is as follows: ?-9-9-?-0-0, 2 Ahau 
13 Pop. This is sufficient to determine the series, and shows the 

above rendering to be correct. 

A B 

Fro. 151 . Part of inscription on the east side of Stela P, Copan. Maudslay. part 4, plate lxxxix. 

Although the drawings in Maudslay \s work arc in most cases of 
unusual excellence, giving details with wonderful accuracy, that of 
this inscription and the one on altar Q (part 4, plate xcili) are not up 
to the usual standard, failing in sonic instances to bring out as 
clearly as might be done some of the minor details. There is some 

22 eth— 04 15 

226 MAYAN CALENDAR SYSTEMS [eth. ann. 22 

doubt as to the value of the face numeral prefixed to the ahau symbol 
(A2, figures 150 and 151), as it is unusual, being in some respects 
unlike any other face numeral that I have observed in the inscrip- 
tions. Seler (Zeitschrift fur Ethnologie, Heft 6, 1899, page 722) 
interprets it 13, and gives as the terminal date 3 Ahau 3 Uayeb. This 
would make the series, omitting the great cycle, 9-9-13-0-0, 3 Ahau 3 
Uayeb; however, the number attached to the month symbol is cer- 
tainly 13. If this series is counted from 4 Ahau 8 Cumhu, it will 
reach 3 Ahau 3 Uayeb in the year 5 Ezanab. The number attached 
to the day Ahau is very uncertain, seeming more like 1 or 3 than 2; 
apparently 1. I have therefore given an exact copy of Maudslay's 
photograph (figure 151), and a carefully made drawing (figure 150), 
nsing Maudslay's and Seler's drawings and the photograph (autotype) 
for this purpose. I am rather inclined to the opinion that Good- 
man's rendering is correct. It seems that Seler has been influenced 
in his determination of the number placed over the Ahau symbol by 
Maudslay's drawing. His interpretation is not justified by the 
photograph, which indicates " 1 Ahau" instead of "3 Ahau," making 
the date 1 Ahau 13 Uo, or 1 Ahau 13 Pop. 

The whole inscription, as well as the inscription on the front and 
back of the same monument, is strange, and, as will be noticed far- 
ther on, shows some of the features of the Chichen Itza inscriptions. 

It is perhaps unnecessary to follow this subject further, as it is 
apparent that the value of the face symbol and other numeral sym- 
bols can be satisfactorily obtained. It appears that Goodman's deter- 
minations, where the data are sufficient, are as a rule correct; though 
there are a few cases, as has been shown, where his rendering is 
doubtful, and some where the series given are largely guess work, 
the data being insufficient. When the number of the great cycle is a 
necessary factor, another question arises, which will be discussed 
farther on. 

Before discussing the numbers of the cycles and great cycles, which 
subject was referred to in my previous paper, I will notice some of 
the secondary numeral series of the Quirigua inscriptions not at 
hand when my previous paper was written. 



Returning to the inscription on the west side- of Stela F (plate lxxi), 
we pass over the first subordinate series (glyph 16), leading on to 6 
Cimi 4 Tzec (glyphs 18 and 19), as this has already been noticed. At 
glyph 25 follows a date, 3 Ahau 3 Mol, but without any recogniz- 
able intermediate numeral series, though there are some numbered 
g ] yphs. Passing on we find at glyph 29 the date 4 Ahau 13 Yax, 
and immediately following (glyph 30 and first half of 31) the numeral 
series 3 days, 13 chuens, 16 ahaus, 1 katun, and following this two 




Fig. 152. Glyph 33, west side 
of Stela F, Quirigua. Mauds- 
lay, part 12, plate XL. 

dates, 12 Caban 5 Kayab (the same as the terminal date of the initial 

series) and 1 Ahau 3 Zip, though the number attached to the day in 

the latter is not the ordinary symbol if intended for 1 (figure 152). 

Counting the series given forward from 3 Ahau 3 Mol and 4 Ahau 

13 Yax brings us to no given date; nor will 

counting back from 12 Caban 5 Kayab reach 

any previous given date. If, however, we 

count back from 1 Ahau 3 Zip, we reach 12 

Caban 5 Kayab, showing that the connection 

is made with the terminal date of the initial 

series, as given by Goodman. It would seem 

from this that the insertion of this date, after 

this second numeral series, is for the purpose 

of showing that the count is to be made from 

this date, as we found in our preceding paper 

to be true in some instances. 

Our next reference is to the* inscription on the east side of Stela F 
(Maudslay's plate XL, part 12). Here the initial series (plate lxxiii), 
as heretofore stated, is 54-9-1 6-10-0-0, 1 Ahau 3 Zip. Goodman, in 
his comment (page 125), says: 

The glyphs that immediately follow are so fantastic and unfamiliar that I can 
make nothing of them until the sign indicating a date to be some score days in 
the 19th katun is reached. The date is 5 Ahau 13 Mol [glyph 24]. As that 
begins the 1st ahau, the number of score days indicated must be 18. Two unin- 
telligible glyphs follow, succeeded by what I believe to be this reading: 3 cycles, 
8 katuns, and 19 ahaus, a reckoning embracing 26 calendar rounds and extending 
360 8-score days into the 13th cycle, to 1 Ahau 13 Yax. the beginning of a 360- 
bissextile count and of a katun also. 

It is somewhat difficult to understand these statements, but I will 
try to explain them, as I desire to offer one or two criticisms. The 

actual interval between 1 Ahau 3 
Zip, the terminal date of the initial 
series, and 5 Ahau 13 Mol (if the 
first following occurrence of this 
date be assumed as the one in- 
tended) will be 18,360 days, or 2 
katuns, 11 ahaus. This will bring 
us to 5 Ahau 13 Mol, the first day 
of ahau number 1 in the 19th katun 
of cycle '•> (as numbered by Good- 
man) — the one now under consider- 
ation. What he means by L8-score 

davs is that the count extends 360 

days into the L9th katun, bringing us to the commencement of the 
second ahau, which, according to his method of numbering, is 1. 

For some unexplained reason, Goodman makes no mention of the 
numeral series between the terminal date of the initial series 1 Ahau 

Fig. 153. Part of inscription on the east side 
of Stela F, Quirigua. Maudslay, part 12, 
plate xl. 

228 MAYAN CALENDAR SYSTEMS [eth. ann. 22 

3 Zip and 5 Ahau 13 Mol. This, unless I am wrong in my interpreta- 
tion, is found in glyphs 21, 22, and 23 (figure 153), as numbered by 
Maudslay. The prefixed numerals with one exception (that prefixed 
to the ahau) are of the ordinary type. However, as the exception, 
which is a face numeral, shows the hand across the lower jaw we must 
assume, according to what has been shown, that it denotes full count 
or naught (0). With this assumption, the series appear to be 3 days, 
11 chuens, ahaus, and 19 katuns, or -19-0-11-3, the number of 
chuens being uncertain; but this series will not connect any pre- 
ceding with anj' following date. Could this have been Goodman's 
reason for omitting notice of the series? 

It is noticeable also that the symbol he interprets 5 in the date 
5 Ahau 13 Mol (glyph 24, figure 153) is precisely the same as the one 
he interprets 1 in the date 1 Ahau 3 Zip in the inscription on the 
west side of this stela (glyph 33). In the next place it is exceedingly 
doubtful, judging from an inspection of the characters, whether his 
supposed series "3 cycles, 8 katuns, and 19 ahaus" can be found in 
the space indicated — that is between glyphs 24 and 29. There is not 
in it, with one exception, a single glyph that in . any way resembles 
any of the forms of time periods he has noticed. The exception is 
the first part of glyph 26, which is like the ordinary form of the chuen 
symbol; but the character over it is like that over Ahau in the date 
he gives as 5 Ahau 13 Mol, elsewhere interpreted as 1. There is a 
numeral, 13, of the ordinary form over the first part of glyph 28, but 
there is no 13 in the series he gives. We take this series, there- 
fore, to be purely imaginary, made up from his tables. According to 
Maudslay's drawing, the month symbol in the following date — 1 Ahau 
13 Yax — is really the symbol for Yaxkin. But an examination of the 
photograph does not bear out the drawing, the glyph being as much 
like the Yax as the Yaxkin symbol. 

According to his statement, this imagined series extends " 360 8-score 
days into the 13th cycle to 1 Ahau 13 Yax." He must, of course, 
allude to the 13th cycle of his 55th great cycle; with this understand- 
ing his count is correct, if he had anything to base it on. 

We turn next to the inscription on the west side of Stela E, the 
drawing of which is shown in Maudslay's plate xxxi, part 12. The 
terminal date of this initial series (see plate lxxii), the number of 
ahaus being corrected from 12 to 13, as already noticed, is 12 Caban 5 
Kayab. The first numeral series which follows is in glyphs 14 and 15 
(figure 154 a), and is 6-13-3 (reversed), equal to 2,423 days. The date 
which follows (glyph 16) is 4 Ahau 13 Yax. The count is correct, as 
2,423 days from 12 Caban 5 Kayab, year 13 Ben, bring us to 4 Ahau 13 
Yax, year 7 Lamat. The next series is found in glyph 18 and, accord- 
ing to the method of reading the chuens and days so far followed — 
that is, counting the number at the left side of the chuen symbol as 
days and that above it as chuens — is, in reverse order, 1-6-14, but 




Goodman, without any explanation, changes it here to 1-14-0. The 
date following (glyphs 19 and 20), is 6 Cimil Tzec. The time given in 
this instance will not reach from one of these dates to the other. As 
Goodman is certainly right in his correction in this instance, if the 
date 4 Ahau 13 Yax be correct, Ave will examine it. The initial 
series of this inscription, including the terminal date, is, when the 
correction noted has been made, precisely the same as that of the 
inscription on the west side of Stela F, and the first following date 
there is the same as the 


second here, 6 Cimi 4 Tzec. 

As the intervening series 
is too short to allow for a 
second return of the latter 
date, it is evident that the 
numeral series must be the 
same. As that of Stela F 
is 13-0-9, then by subtract- 
ing, in the inscription on 
Stela E, the 6-13-3 extend- 
ing from 12 Caban 5 Kayab 
to 4 Ahau 13 Yax, from this 
series (13-0-0) the remain- 
der, 6-14-6, must give the 
lapse of time from 4 Ahau 
13 Yax to 6 Cimi 4 Tzec, the 
third date, and calculation 
shows that it does. There- 
fore the correction from 
1-6-14 to 1-14-6, and the 
1 to 6, giving 6-14-6, may 
be accepted as justifiable 
if the date 4 Ahau 13 Yax 
be correct. At any rate, it 
is certain that this change 
is correct or that an equiv- 
alent change in the preced- 
ing series 6-13-3, must be 
made and the date altered to suit. I am therefore inclined to accept 
the correction made by Goodman. 

Following the last date at glyphs 21 and 22 (figure 155) is the series 
15 days 18 chuens 1 ahau 1 katun, or in reverse order 1-1-18-15. 
The numbers are distinct and of the ordinary type, and are given 
correctly, as is shown by inspection both of the photograph and draw- 
ing. That there is an error here (18 chuens being full count) seems 
apparent, unless the number a1 the led side of the chuen symbol 
refers to chuens and that above to days, which can not be accepted 

Fig. 154. Part of inscription on the west side of Stela E, 
Quirigua. Maudslay, part 1:>, plate xxxi. 



[ETH. ANN. 22 

without proof. Goodman reads " 1-1-16-15," but the number over the 
symbol is 18 and not 16. The two outer of the three units are cer- 
tainly balls, and not rings or semicircles. This series is followed at 
glyphs 23 and 24 (figure 155) by the date 11 Imix 19 Muan, and whether 


we count 18 or 16 chuens or consider the 15 as chuens and the 18 as 
days, it fails to connect the preceding with the following date. Before 
attempting to find the solution of the difficulty we will pass on to the 
next series and date and count back. 

Passing on to glyphs 27 and 28 (figure 156) we find the series 8-19-4, 
followed (glyph 29) by the date 13 Ahau 18 Cumhu, and this is fol- 
lowed immediately (glyph 30) by the symbol for 17 katuns, apparently 
inserted, as it is followed by no date, to show that the date just 
preceding it is in the 17th katun, or that 17 katuns have been passed 
over from the commencement of the cycle, most likely the latter. 

As Goodman does not 
discuss this series, 
although he mentions it, 
I give my own explana- 
tion. That there is an 
error here, if the number 
over the chuen symbol is 
intended to indicate chu- 
ens, as there are but 18 
chuens in an ahau, is 
apparent. Let us try 
the count with the day 
and chuen numbers re- 
versed — that is, on the 
supposition that the se- 
ries should read 8-4-19. 
This equals 2,979 days, 
which number counted 
backward from 13 Ahau 18 Cumhu brings us to 11 Imix 19 Muan, 
which apparently justifies the change and proves the date ''11 Imix 
19 Muan " to be correct. 

Turning to Goodman's "Archaic Chronological Calendar," to the 9th 
cycle of his 54th great cycle, in which the series of this inscription 
are located, we find that 13 Ahau 18 Cumhu is the first day of the 17th 
katun according to his method of numbering. However, it must be 
remembered that he begins the count of katuns with 20, following 
with 1, 2, etc., up to 19; therefore 13 Ahau 18 Cumhu is really the 
first day of the 18th katun, 17 entire katuns having been passed over 
from the initial date of the inscription (8 Ahau 13 Ceh, the first day of 
the 9th cycle). This verifies our conclusion as to the signification of 
the symbol for 17 katuns in glyph 30. 

Fig. 155. Part of inscription on the west side of Stela E, 
Quirigua. Maudslay, part 12, plate xxxi. 


For the purpose of determining the third minor series given in the 
inscription as 1-1-18-15, followed by 11 Imix 19 Muan, we will count 
from the initial date of the inscription, placing side by side the series 
as given in the inscription and as corrected. 

As given As corrected 

Initial 9_14-12-4-17 9-14-13-4-17 

Second 6-13-3 6-13-3 

Third 1-6-14 6-14-6 

Fourth _ . .. 1-1-18-15 

Fifth 8-19-4 8-4-19 

9_16-ll-6-13 9-15-15-1-5 

If we subtract 0-15-15-1-5, the sum of the right column (omitting 
the 4th series), from 9-17-0-0-0, or, omitting the cycles, 15-15-1-5 
from 17-0-0-0 (17 katuns), the remainder is 1-4-10-15, or 1 katun 4 



Fig. 150. Part of inscription on the west side of Stela E, Quirigua. Maudslay, part 12, 

plate xxxi. 

ahaus 10 chuens and 15 days. This, if the preceding corrections are 
justified, should be the 4th series, and should connect (counting for- 
ward) the dates G Cimi 4 Tzec and 11 Imix 1!> Muan, and calculation 
shows that it does. The 4th series should therefore be 1-4-10-15, or 
8,075 days. 

It will be seen from our examination of this inscription that some 
correction has been made in the 1st, 3rd, and 41 li series, and that the 
day and chuen numbers have been reversed in the 5th. It musl be 
admitted that this docs not present a very favorable showing for the 
theory, yet I am convinced thai the corrections in this instance are 


justified; but a single variation is possible (that of the 3rd date) 
which would involve greater changes than those which have been 
made. That the number at the left of the chuen symbol sometimes 
denotes chuens and the one over the top sometimes denotes days is 
mentioned by Maudslay, yet it is very unusual and is probably due to 
carelessness. There is evidence of carelessness in this inscription in 
tho writing of 18 and 19 chuens, and in giving 12 ahaus in the initial 
series instead of 13, as it evidently should be. 

The next inscription referred to is that on the east side of Stela E, the 
drawing of which is shown in Maudslay's plate xxxn, part 12 (our plate 
lxxviii). The initial series is 54-9-17-0-0-0, 13 Ahau 18 Cumhu, 
Goodman does not mention this inscription. It ends precisely where 
the preceding inscription ended. Although there are distinct dates 
scattered through it, and what appear to be partial series, I am unable 
to determine the latter from the unusual symbols of which they are 
formed, if they are present. The inscription appears to end, so far 
as dates are concerned, with 13 Ahau 18 Cumhu, the same as the 
terminal date of the initial series, which does not occur again in 
Goodman's tables until the beginning of the 9th ahau 4th katun 12th 
cycle is reached. This gives a lapse of 2-7-9-0-0 from the terminal 
date of the initial series. As nothing further in regard to the series 
can be learned from this inscription, we turn to that on Stela A, 
Maudslay's plate vn, part 11. 

The initial series on Stela A is, as has been shown, 54-9-17-5-0-0, 
6 Ahau 13 Kayab. Immediately following the month symbol of the 
date (glyph 16) is the symbol for 6 Ahau. This, I believe, is to show 
that the preceding date is the beginning of the 6th ahau, and so it is 
if we count the ahaus 1, 2, 3, etc., from the commencement of the 
katun, instead of 20, 1, 2, 3, etc., as Goodman counts them. It is my 
belief that the numbers expressed in the series denote, at least as a 
general rule, completed periods and not incomplete ones. Take, for 
example, the numbers in the initial series in this inscription, omitting 
the great C3 T cle — 9-17-5-0-0, that is, 9 cycles, 17 katuns, 5 ahaus, 
chuens, days. This may be read just as I have given it here, or 
as follows: The 5th ahau of the 17th katun of the 9th cycle. If it 
should be read as I have given it, it shows that Goodman's method of 
counting — beginning that of the cycles with 13 following with 1, 2, 3, 
etc., that of the katuns and ahaus with 20, and following with 1, 2, 
3, etc. — is erroneous. If we read 9 cycles, 17 katuns, and 5 ahaus, 
the meaning is that 9 full cycles, plus 17 katuns, plus 5 ahaus must 
be counted to make the sum of the days between the preceding and 
following date, and this is in fact the method Goodman uses, and 
which must be used in making the calculation. On the other hand, 
according to his system, the series 9-17-5-0-0 would indicate that the 
date sought is the 1st day of the 5th ahau of the 17th katun of the 






9th cycle, but the symbol 6 Ahau (glyph 16) denotes, if we have cor- 
rectly interpreted it, that 6 Ahau 13 Kayab is the first day of the 6th 
ahau; nevertheless, Goodman's method of counting gives the correct 
result. Attention will again be called to the subject further on. 

Returning to our inscription, we find in the 20th glyph the brief 
series 19 ahaus followed by the date 6 Ahau 13 Chen or 13 Zac, but 
the series does not connect the dates. There are no other recogniza- 
ble series in the inscription. 

The inscription on the west side of Stela C — the drawing of which 
is shown in Maudslay's plate 19, part 11 (our figure 147) — has, as 
heretofore stated, the initial series 54-9-1-0-0-0, 6 Ahau 13 Yaxkin. 
Following this date, at glyphs 16 and 17, is the numeral series 17-5-0-0, 
that is, 17 katuns, 5 ahaus, chuens, days, though in the usual 
reverse order of days, chuens, ahaus, katuns. This is in turn fol- 
lowed by the date 6 Ahau 13 Kayab. If we count this series as 16 
katuns and 5 ahaus, it will exactly express the lapse of time from 6 
Ahau 13 Yaxkin, the preceding date, to 6 Ahau 13 Kayab, the date 
which follows. But turning to Goodman's "Archaic Chronological 
Calendar," 54th great cycle, we find that the latter date, according to 
his numbering, is the 5th ahau of the 17th katun of the 9th cycle. 
Shall we accept this as the proper reading, or shall we conclude that 
there is an error in the number of katuns? 6 Ahau 13 Yaxkin is 
the first day of the 1st katun of the 9th cycle, according to Goodman's 
method of counting (though the 2nd, in fact, if the count began with 
1), and 6 Ahau 13 Kayab is the first day of the 5th ahau, as Goodman 
counts (6th in fact), of the 17th (18th) katun. Counting from one 
date to the other gives just 16 katuns 5 ahaus, as the following sub- 
traction shows: 

9- 1-0-0-0 


It is proper to bear in mind that by Goodman's method of number- 
ing, the number given always expresses the number completed; thus, 
as he begins with 13 in numbering the cycles, his 1st cycle is in reality 
the second, one cycle having been completed and the 2nd entered 
upon. I am therefore disposed to correct 17 katuns in the series just 
examined to 16. 

As these are the only series of the Quirigua inscriptions to which it 
is desirable to call attention at present, the next subject of examina- 
tion is the great-cycle symbols, but in order to enter upon this intelli- 
gently it is necessary to discuss some points of Goodman's system not 
fully examined in my previous paper. In doing this it will be neces- 
sary to go to the very base of his system. 

234 MAYAN CALENDAR SYSTEMS [eth. ann. 22 


The theory that Goodman has adopted, so far as it relates to the 
seal? of units or time periods, as he terms them, may be expressed in 
the following series, the day being the primary unit: 

Day 1 day 

20 days make 1 chuen 20 days 

18 clmens make 1 ahau 360 days 

20 ahaus make 1 katun 7,200 days 

20 katuns make 1 cycle 144,000 days 

13 cycles make 1 great cycle 1,872,000 days 

73 great cycles make 1 grand era 136,656,000 days 

This scheme is, as was explained in my previous paper, precisely the 
same as that generally accepted, so far as the numbers are concerned, 
until, in ascending the scale, the number of cycles, or units of the 5th 
order, forming a great cycle, or unit of the next higher order, is reached. 
At this point Goodman abandons the vigesimal system and introduces 
in one step 13 and in the other 73 as multipliers — numbers which are 
absolutely necessary to his theory; for if either be dropped, his theory 
falls with it. If these supposed time periods are, as I contend, noth- 
ing more than orders of units in the system of numeration, then we 
must assume that the vigesimal system was followed. To this point 
attention is directed, and although it is discussed somewhat at length 
in my previous paper, there is other evidence bearing on the question, 
which will be introduced here. It was shown there that one series 
in the Dresden codex recognizes 20 cycles to the great cycle (I shall 
continue to use these terms merely for convenience, to indicate the 
orders of units). A more careful study of that codex shows that there 
are other series which also furnish conclusive evidence on this point. 

The theory, therefore, which I shall attempt to show is the correct 
one is that in both the Dresden codex and the inscriptions the viges- 
imal system was maintained throughout, except only in the second 
step; not only that 20 ahaus make 1 katun and 20 katuns make 1 
cycle, but also that 20 cycles make 1 great cycle and 20 great cycles 
1 next higher step, should the count extend so far. 

Before we consider the examples which are to be introduced as. evi- 
dence in support of this theoiy, it will be best, in order to see more 
clearly the bearing and the force of this evidence on the question, to 
present an explanation of the order of succession of the great cycles 
when the vigesimal system is followed, that is, when 20 cycles are 
counted to the great cycle. 

As the day Ahau is found to be the first day of several, in fact most, 
of the initial series of the inscriptions, and is that adopted by Good- 
man as the beginning of his grand era, as also of his great cycles, I, 
for the present, assume it as the initial day of the latter periods. 

According to his scheme of counting 13 cycles to each of these 


periods, they all begin with the day 4 Ahau. If the first day of the 
ahaus is Ahau, then it is certain that the first day of each of the 
higher periods will be Ahau, though we count 13 or 20 cycles to the 
great cycle. As the days of the calendar are numbered 1, 2, 3, etc., 
up to 13, the count then beginning again with 1, and this numbering 
is continued in regular order, and as Ahau will return only every 201 h 
day it is apparent that it will receive different numbers. If the days 
are written out in regular succession and the series is made of suffi- 
cient length, it will be found, if we select a 13 Ahau and begin our 
count with it and count 360 daj T s (1 ahau) to each step, that the num- 
bers attached to the days (which will of course be Ahaus) will come (the 
count being forward) in the following order: 13, 9, 5, 1, 10, 6, 2, 11, 7, 
3, 12, 8, 4, 13, 9, 5, etc., this order being maintained wherever in the 
series we may begin. 

As it takes 20 ahaus or units of the 3rd order to make one of the 
4th, it follows that if the day numbers are written out in succession 
in the order above stated, the first days of the katuns will be those of 
the 20th ahaus, their numbers will therefore come in the following 
order: 11, 9, 7, 5, 3, 1, 12, 10, 8, 6, 4, 2, 13, 11, 9, 7, etc., the order 
remaining the same regardless of the point at which the count begins. 
As 20 katuns make 1 cycle, the numbers of the first days of the 
cycles will be the same as those of the 20th katuns, and will be as fol- 
lows: 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 13, 12, etc. The beginning 
point in these series is arbitrary. 

It may also be shown by simple calculation that the order of the 
day numbers of the first days of the higher periods or orders of units 
will be as given above. As the numbers of the first days of the ahaus 
vary successively by 4, if we multiply 4 by 20 (20 ahaus being required 
to make a katun) and divide by 13, the remainder is 2; hence, if the 
first day of a given katun is 9, the first day of the one which follows 
will be 7 Ahau, the difference being subtracted if counting forward, 
and added if counting backward. When the number of the day is 
less than 3 we add 13, and then subtract in counting forward, and in 
counting backward subtract 13 when the sum is greater than this 
number. As it takes 400 ahaus to make 1 cycle, we multiply the 
difference, 4, by this number, and divide the product by 13. This 
leaves a remainder of 1, hence we subtract 1 from the number of the 
first day of a given cycle to find the first of that which follows, or 
add 1 to find the first of that which precedes. 

As, according to Goodman's theory, 13 cycles make a great cyel<\ 
then it requires 20x20x13 alums to make 1 great cycle. We mul- 
tiply 4 by 20x20x13 (or 5,200) and divide by 13. This leaves no 
remainder, and hence, according to this scheme, the day numbers of 
the first day of all the great cycles will be the same, and so Goodman 
gives them in his "Perpetual Chronological Calendar." Here the 
question of number arises. Is it 1 Ahau, 2 Ahau, or 3 Ahau, etc., to 

236 MAYAN CALENDAK SYSTEMS [eth. ann. 22 

13 Ahau? Goodman says 4 Ahau. He bases this, doubtless, on the 
fact that many of the initial series of the inscriptions have as their 
first day 4 Ahau 8 Cumhu, which he assumes, apparently correctly, is 
the first day of a great cycle. It is apparent, following his method of 
numbering, that if one great cycle begins with 4 Ahau, all the rest do. 
As yet we have not introduced the year as a factor, but before this 
is done attention is called to the result of following the vigesimal sys- 
tem in counting the higher orders of units, or time periods, as Good- 
man considers them. According to this system, which, as I have 
stated, prevails in the Dresden codex, not only does it take 20 ahaus 
to make 1 katun and 20 katuns to make 1 cycle, but also 20 cycles 
to make 1 great cycle. The order in which the numbers of the initial 
days of the ahaus, katuns, and cycles follow one another will be the 
same in the one scheme as in the other and as already given. The 
difference between the two theories appears in the numbers of the 
initial days of the great cycles. Following the method of calculation 
indicated, we multiply 4 by 20x20x20 (or 8,000) and divide by 13. 
This gives a remainder of 7. The order of the numbers is therefore 
13, 6, 12, 5, 11, 4, 10, 3, 9, 2, 8, 1, 7, 13, 6, 12, etc., anfl this is found 
to be correct by the absolute test of writing out the numbers of the 
first days of the cycles in proper order and taking every 20th one. 
The initial dates of a sufficient number to cover all probable require- 
ments are given here, 4 Ahau 8 Cumhu being adopted as the basis or 
check point from which to count forward and backward. In this 
calculation we must bring into the problem the year factor. 

Initial days of the great cycles, following the vigesimal system 

1 5 Ahau 8 Mnan, year 4 Ben 

> 2 11 Ahau 13 Zotz, year 4 Lamat 

3 4 Ahau 3 Ceh, year 3 Ezanab 

4 10 Ahau 8 Pop, year 3 Ben 

5 3 Ahau 18 Mol, year 2 Akbal 

6 9 Ahau 8 Pax, year 1 Ben 

7 2 Ahau 13 Tzec, year 1 Lamat 

8 8 Ahau 3 Mac, year 13 Ezanab 

9 1 Ahau 8 Uo, year 13 Ben 

10 7 Ahau 18 Chen, year 12 Akbal 

11 13 Ahau 8 Kayab. year 11 Ben 

12 6 Ahau 13 Xul, year 11 Lamat 

13 12 Ahau 3 Kankin, year 10 Ezanab 

14 5 Ahau 8 Zip, year 10 Ben 

15 11 Ahau 18 Yax, year 9 Akbal 

16 4 Ahau S Cumhu, year 8 Ben 

17 10 Ahau 13 Yaxkin, year 8 Lamat 

18 3 Ahau 3 Muan, year 7 Ezanab 

19 9 Ahau 8 Zotz, year 7 Ben 

20 2 Ahau 18 Zac, year 6 Akbal 

As no larger number of great cycles has been recorded than 14, in 
one of the Copan inscriptions, 6 being the highest given in the Dres- 


den codex, the initial dates given will probably suffice for all require- 
ments. But this supposition rests on the theory that the range 
counting by great cycles, is not more than 14 from 4 Ahau 8 Cumhu. 
Our numbering (left column) is, of course, purely arbitrary, given 
merely for convenience of reference, the great cycles being, on the 
theory I have presented, in precisely the same relation to the next 
higher order of units — provided the Mayan count extended so far — as 
the cycles to the great cycles, the katuns to the cycles, etc. In other 
words, when, in counting, 20 cycles are completed, one great cycle is 
completed and the count passes into the 2nd; and when this is com- 
pleted we pass into the 3rd, etc., in precisely the same manner that 
we pass in our decimal system from one decimal to the next higher. 

Our next step is to test the theory advanced by appeal to the high 
series which reach to the great cycles, beginning with those of the 
Dresden codex. These are found on plates lxi, lxii, and lxix. As 
the determination of the point in question is of vital importance, the 
details of the demonstration will be given somewhat fully. 

Taking first plate lxii of the codex (our plate lxxix), we observe 
four numeral series running upward in the folds of two serpent 
figures, two of these series being in black numerals of the ordinary 
form, and two in red, also of the ordinary form. The two series in 
the left serpent (one black and the other red) are as follows reading 
from the top down: 

Red 4-6-11-10-7- 2, 3 Cimi 14 Kayab 

Black _. 4-6- 7-12-4-10, 3 Ix 7 Pax (?) 

That is to say, the red series is 4 great cycles, 6 cycles, 11 katuns, 
10 ahaus, 7 chuens, 2 days, to 3 Cimi 14 Kayab. The symbols of the 
dates as we give them are seemingly reversed as compared with their 
positions on the plate, but the zigzag order of the series must be borne 
in mind. The symbol of the month Pax is somewhat unusual. 

The red series changed into days is as follows : 


4 great cycles (of 20 cycles each) 11, 520, 000 

6 cycles 864, 000 

11 katuns 79,200 

10 ahaus 3,600 

7 chuens 140 

2 days 2 

Total amount 12, 466. 94i 

Subtract 655 calendar rounds 12, 450, 880 

Remainder 16, 062 

Using this remainder and counting forward from U Kan 12 Kayab 
(year 3 Ben) — the date standing over the head of the figure seated on 
the serpent — we reach 3 Cimi 14 Kayab, year 8 Ben, the date standing 

We have positive evidence, therefore, that in this instance 9 Kar 

238 MAYAN CALENDAR SYSTEMS [eth. ann. 22 

12 Kayab is the initial day of a great cycle and that 20 cycles are 
counted to the great cycle, since the number 11,520,000 is obtained 
as follows : 

1 cycle 144, 000 days 

Multiplied by _ A 20 

1 great cycle 2, 880, 000 days 

Multiplied by 4 

4 great cycles 11, 520, 000 days 

If we follow Goodman's method and count only 13 cycles to each 
great cycle, 4 of the latter, together with the minor periods of the 
series as given above, will amount to 8,432,942 days. Subtract 444 
calendar rounds, and there remain 5,822 days, which, counted from 9 
Kan 12 Kayab, bring us to 7 Cimi 14 Pax. This is not correct as to 
the number of the day or as to the month. The same day should 
be reached, for the number of cycles is the only thing in the series 

We take next the black series of the same pair, to wit, 4-G-7-12-4- 
10, 3 Ix 7 Pax. This changed into d°.ys is as follows: 


4 great cycles (of 20 cycles each) 11, 520, 000 

6 cycles 864, 000 

7 katuns . 50, 400 

12 ahaus 4, 320 

4 chuens 80 

10 days 10 

Total . 12, 438, 810 

Subtract 655 calendar rounds 12, 431 , 900 

< Remainder 6, 910 

Using this remainder and counting forward from 9 Kan 12 Kayab, 
year 3 Ben, the same initial date as before used, we reach 3 Ix 7 Pax, 
year 9 Lamat. This is correct. 

The series in the folds of the right serpent (same plate as the pre- 
ceding) are as follows: 

Black 4-6-9-15-12-19, 13 Akbal 1 Kankin 

Red 4-6-1- 9-15- 0, 3 Kan 16 (?) Uo 

Changing the red series into days, we have the following result : 


4 great cycles (of 20 cycles each) 11, 520, 000 

6 cycles 864,000 

lkatun , 7,200 

9 ahaus 3,240 

15 chuens . 300 

Total 12,394,740 

Subtract 653 calendar rounds 12, 393, 940 

Remainder . 800 





Using this remainder and counting forward from 9 Kan 12 Kayab 
(same initial date as before), we reach 3 Kan 17 Ho, year 6 Lamat. 
This is correct, as it gives the date below, except as to the day of the 
month — which is given as 16 Uo in the original, but should be 17 Uo, 
as Kan is never the 16th day of the month. What is meant by the 
calendar rounds and the reason for subtracting them was fully 
explained above and in my previous paper. 

The black series of the same pair changed into days gives the fol- 
lowing numbers: 


4 great cycles (of 20 cycles each) 11, 520, 000 

6 cycles 864, 000 

9 katuns 64, 800 

15 ahaus 5,400 

12 chuens 240 

19 days 19 

Total : 12, 454, 459 

Subtract 656 calendar rounds 12, 450, 880 

Remainder _■ _ _ . 3, 579 

Counting forward this number of days from 9 Kan 12 Kayab, year 
3 Ben, we reach 13 Akbal 1 Kankin, year 13 Akbal. This also is 

The next series noticed is the one consisting of black numerals in 
the folds of the serpent on plate lxix of the Dresden codex (our 
plate lxxx). This is as follows : 4-5-19-13-12-8, 4 Eb ? (month) ; the 
month symbol is obliterated. As the black and red are not zigzagged 
in this instance, the date belonging to the black series stands imme- 
diately under it. Changed into days, the series gives the following 
result : 


4 great cycles (of 20 cycles each) 11, 520, 000 

5 cycles 720, 000 

19katnns .... 136,800 

13 ahaus 4, 680 

1 2 chuens 240 

8 days 8 

Total 12,381,728 

Subtract 652 calendar rounds 12, 374, 960 

Remainder 6, 768 

In this instance, as on plate lxii of the codex, the date 9 Kan 12 
Kayab stands above the serpent. Counting forward 6,768 days from 
this date, we are brought to 4 Eb 5 Chen, year 9 Lamat, which agrees 
with the unobliterated part of the date given below. 

We have, therefore, in the data presented positive proof that in 
five instances in the Dresden codex the day 9 Kan 12 Kayab is the 
first day of a great cycle, and that twenty cycles are counted to one 

240 MAYAN CALENDAR SYSTEMS [eth. ann.22 

great cycle. In these instances 9 Kan 12 Kayab is the initial clay of 
the first or more remote of the four great cycles counted in the series 
which have been noticed. The four here, however, has no reference 
to the numbers applied to the high periods, if, in fact, any were 
applied, but is merely the number of one of the orders of units used 
in counting, just as we say " 4 thousands, 5 millions," etc. However, 
the idea intended to be set forth here will be more fully explained 
farther on. 

In order to show that 9 Kan, as used in the series examined, is the 
initial day of the most remote of the four great cycles of these series, 
the following proof is presented. 

If we arrange the last-mentioned series perpendicularly in ascend- 
ing order, as in the original, except that we separate the great cycles, 
it will stand as follows: 

4th great cycle (completed) 

3rd great cycle (completed) 

2nd great cycle (completed) 

1st great cycle (completed) 

5 cycles 

19 katuns , 

13 ahaus 

12 chuens 

8 days 

The reader must keep in mind all the way through that, although 
Goodman's terms are used, they are to be understood as representing 
merely orders of units. Hence, 4th great cycle, 3rd great cycle, etc., 
are intended to convey the same idea that is conveyed by "4th mil- 
lion, 3rd million," etc. These terms are used merely as convenient 
designations in numeration. Each and every series in the inscrip- 
tions and codices signifies nothing more nor less than so many days, 
the day being the unit. 

Our separation of the great cycles is therefore nothing more than 
separating the millions and lower denominations in the expression 
" 4,234,600," just as has been done above. The object of this separa- 
tion is to ascertain the beginning day of each of these numbers which 
Goodman calls time periods, as this forms a check on our calculations. 
For example, if I assert that 4,000 days from Thursday, Januaiy 1, 
1889, will reach Saturday, December 18, 1899, by counting 1,000 days 
we reach a certain date, and 1,000 more a certain other date, etc. 
If the fourth 1,000 brings us to the same date as counting at once 
4,000, we thereby check the one calculation by the other. The sepa- 
ration is to be understood as signifying nothing more than this, and 
not as implying real time periods of a chronological system. 

If we can ascertain the first day of the first of these great cycles, 
and count forward from the date so obtained, one by one, 4 great 





cycles, 5 cycles, 19 katuns, 13 ahaus, 12 chueus, and 8 days, we should, 
if my theory be true, reach the same date (4 Eb 5 Chen, year 9 Lamat) 
as b} T counting the whole series, thus obtaining a check on our 


Multiply 1 cycle 144,000 

by 20 20 

1 great cycle of 20 cycles 2, 880, 000 

Subtract 151 calendar rounds 2, 865, 980 

Remainder 14, 020 

Counting forward this number of days from 9 Kan 12 Kayab, year 
3 Ben, we reach 2 Kan 17 Xul, year 3 Lamat. This should be the 
initial day of the 3rd great cycle, as numbered above. Counting 
forward 14,020 days from 2 Kan 17 Xul, year 3 Lamat, brings us to 
8 Kan 7 Kankin, year 2 Ezanab. This should be the first day of the 
2nd great cycle, as numbered above. Counting forward 14,020 days 
from the latter date (8 Kan 7 Kankin, year 2 Ezanab), we reach 1 Kan 
2 Zip, 3'ear 2 Ben. This should be the first day of the 1st great 
cycle, as numbered above, and with the subordinate periods gives the 
series 1-5-19-13-12-8, or 1 great cycle, 5 cycles, 19 katuns, 13 ahaus, 
12 chuens, 8 days. Counting forward from 1 Kan 12 Zip, year 2 Ben, 
should bring us to 4 Eb 5 Chen, year 9 Lamat, the date obtained by 
counting the entire series from 9 Kan 12 Kayab, year 3 Ben. 

In order to test it we make the calculation; reduced to days, the 
result is as follows : 


1 great cycle (of 20 cycles) 2, 880, 000 

5 cycles , 720, 000 

19katuns 136,800 

13 ahaus 4, 680 

12 chuens 240 

8 davs 8 

Total 3,741,728 

Subtract 197 calendar rounds 3, 739, 060 

Remainder 2, 668 

Counting forward this number of days from 1 Kan 12 Zip, year 2 
Ben, we reach 4 Eb 5 Chen, year 9 Lamat, the date at the bottom of 
the series, and the same as that obtained by using the entire series 
and counting from 9 Kan 12 Kayab. 

As a further test, we count forward 14,020 days from 1 Kan 12 Zip, 
year 2 Ben, and reach 7 Kan 2 Zac, year 1 Akbal. This should be the 
first day of the incomplete great cycle in which the minor periods 
fall. Therefore, by taking the sum of these periods and counting 
forward from this date, we should reach 4 Eb 5 Chen, year 9 Lamat. 

22 eth— 04 16 



[ETH. ANN. 22 

Reducing these periods (5 cycles, 10 katuns, 13 ahaus, 12 chuens, 8 
days) to days, we get the following result: 


5 cycles 1 720, 000 

19 katuns 136, 800 

. 13 ahaus .* 4, 680 

12 chuens 240 

8 days 8 

Total 861 , 728 

Subtracting 45 calendar rounds 854, 100 



Counting forward 7,628 days from 7 Kan 2 Zac, 3 7 ear 1 Akbal, we 
reach 4 Eb 5 Chen, year 9 Lamat, which is the proper date. 

The demonstration therefore seems to be complete that Kan, in 
the cases referred to, is the first day of each of the great cycles. It 
is also important to notice that the numbers of these Kans follow one 
another in precisely the same order as do those of the Ahaus when 
20 cycles are counted to the great cycle (see page 236) to wit: 9, 2, 8, 
1, 7, and, if the series is continued b} 7 calculation, 13, 6, 12, 5, 11, 4, 
10, 3, 9, 2, etc. 

If we arrange these first days of the great cycles in the order in 
which the} 7 come, adding the days of the month on which the} 7 fall, 
they will be as follows — the numbering (column at the left) being, of 
course, purely arbitrary: 

1 2 Kan 17 Cumhu, year 10 Lamat 

2 8 Kan 2 Mol, 

3 1 Kan 12 Muan, 

4 7Kanl7Zotz 

5 13 Kan 7 Ceh, 

6 6 Kan 12 Pop, 

7 12 Kan 2 Chen, 

8 5 Kan 12 Pax, 

9 11 Kanl7Tzec, 

10 4 Kan 7 Mac, 

11 10 Kan 12 Uo, 

12 3 Kan 2 Yax, 

13 9 Kan 12 Kayab, 

14_...__ 2Kanl7Xul, 

15 8 Kan 7 Kankir 

16. 1 Kan 12 Zip, 

17 7 Kan 2 Zac, 

18 13 Kan 12 Cumhu, year 13 Ben 

19 6 Kan 17 Yaxkin, year 13 Lamat 

20 12 Kan 7 Muan, year 12 Ezanab 

This is calculated from 9 Kan 12 Kayab as a basis, because we have 
found it to be such for some of the series of the Dresden codex. 

In order to add proof to our explanation 'and calculation of the 
series in the serpent figures of plate lxii of the codex, I show the result 


10 Akbal 


9 Ben 


9 Lamat 


8 Ezanab 


8 Ben 


7 Akbal 


6 Ben 


6 Lamat 


5 Ezanab 


5 Ben 


4 Akbal 


3 Ben 


3 Lamat 


2 Ezanab 


2 Ben 


1 Akbal 


of calculating the differences between the series and passing from 
one of the final dates to the other. I had tried this before, but, not 
allowing for the zigzag course of the series, I failed to get the dates at 
the bottom in right relation to the series. 

Take first the series in the right-hand serpent, as follows : 

Black . 4-6-9-15-12-19, 13 Akbal 1 Kankin 

Red 4-6-1-9-15-0, 3 Kan 16 (17) Uo 

Difference 8-5-15-19 

This difference, counted forward from 3 Kan 17 Uo (the 10 being an 
error), should reach 13 Akbal 1 Kankin. 

Reducing to days, we have the following result : 


8 katuns 57, 600 

5 ahaus 1 , 800 

15 chnens 300 

19 days ._ 19 

Total 59, 719 

Subtract 3 calendar rounds 56, 940 

Remainder 2, 779 

Using this remainder and counting forward from 3 Kan 17 Uo, year 
6 Lamat, we reach 13 Akbal 1 Kankin, year 13 Akbal. This is cor- 
rect, and proves that we should read 17 Uo instead of 10. 

The two series in the other (left-hand) serpent are as follows: 

Red .. 4-6-11-10-7- 2, 3 Cimi 14 Kayab 

Black 4-6- 7-12-4-10, 3 Ix 7 Pax 

Remainder 3-18-2-12 

This remainder, counted forward from 3 Ix 7 Pax, which is the date 
belonging to the black series, will bring us to 3 Cimi 14 Kayab, which 
is the date belonging to the red series. 

The relation between the pairs of the two serpents is between the 
like colors. For example, by using the difference between the red 
series of the right serpent and that of the left, and counting forward 
from 3 Kan 17 Uo, we reach 3 Cimi 14 Kayab. By using the differ- 
ence between the black series, and counting forward from 3 Ix 7 Pax, 
we reach 13 Akbal 1 Kankin. These results serve to confirm the 
results of the calculations when the entire series is taken into the 

There are five other high series in the Dresden codex, to which I 
have not- as yet alluded — four in the serpent figures on plate LXI, and 
the red series in the serpent on plate lxix. The reason for passing 
over them temporarily is that some of them require correction, and 
others present difficulties to successful calculation and satisfactory 
interpretation which I have not as yet been able to overcome. As the 
object in view is to discover the truth and not merely to support a 

244 MAYAN CALENDAR SYSTEMS [eth. ann. 22 

theory, it is proper that these difficulties should be explained to the 
reader that he may judge whether they have any bearing on the ques- 
tion under discussion. 

The first of these series to which reference will be made are the 
black and red in the left serpent on plate lxi of the codex (our plate 
lxxxi). These, as they stand on the plate, are as follows: 

Black 4_6-l4-13-15-l , 3 Chicchan 13 Pax 

Red 4-6- 0-11- 3-?, 3 Chicchan 18 Xul 

In this instance, as on plate lxii, the dates under the series are 
here seemingly reversed by the zigzag arrangement of the series — a 
fact which is to be borne in mind; therefore, that which is apparently 
under the black belongs to the red. The last (lowest) number of the 
red series denoting days is obliterated, but calculation soon makes it 
apparent that it was 1. The initial date here is the same as that of 
the other series of this codex heretofore referred to, to wit, 9 Kan 

12 Kayab, which stands in the text above the serpent. 
Calculating the series as the} 7 stand in the original, counting from 

the initial date (9 Kan 12 Kayab), we find, whether we assume 20 or 

13 cycles to the great cycle, that neither of the dates standing below 
will be reached. The proxDer day, and even the day of the month, may 
be reached, but not the full date as given. Counting 20 cycles to the 
great cycle, we are brought by the black series to 1 Chicchan 18 Chen, 
year 6 Lamat; the red series (adding one day) brings us to 5 Chic- 
chan 13 Mac, year 13 Ben. The result in both cases is wrong. Count- 
ing 13 cycles to the great cycle in the black, we reach 3 Chicchan 13 
Kayab, year 9 Ben; and the red series brings us to 7 Chicchan 3 Zip, 
year 4 Akbal. Both results are wrong, though the first is apparently 
within one month of being correct — the day, day number, and day of 
the month being right. However, the two dates are in reality 32 years 
apart. We might assume the number of months (chuens) to be 14, 
instead of 15 as given in the original, if this would bring both series 
in harmony; or we might change the month from Pax to Kayab, if this 
would meet the difficulty throughout. The two series, black and red, 
are evidently related, and the difference between them must connect 
the dates reached by counting each series from the initial date (9 Kan 
12 Kayab). The difference in this case, 13 cycles being counted to 
the great cycle, brings the red series to 7 Chicchan 3 Zip, year 4 
Akbal, which is wrong. 

With seeming inconsistency, T propose a correction more radical 
than either of those suggested above. I believe the aboriginal artist 
by inadvertency made an exchange between the black and red series 
in the ahaus and chuens, and that, instead of being as given above, 
they should be as follows: 

Black ... 4-6-14-11- 3-1, 3 Chicchan 13 Pax 

Red 4-6- 0-13-15-1, 3 Chicchan 18 Xul 



y« — - 

: 8» 




The series evidently requires that the days of the terminal dates 
shall each be 3 Chicchan. 

Counting forward from 9 Kan 12 Kayab, year 3 Ben, the amount of 
the black series (equaling 12,488,821 days), we reach 3 Chicchan 13 
Pax, year 3 Ben; and counting from the same initial date the red 
series (equaling 12,388,081 days), we reach 3 Chicchan 18 Xul, year 
3 Lamat. Both results are correct, and counting from 3 Chicchan 18 
Xul, year 3 Lamat, the difference between the two series as thus cor- 
rected (equaling 99,840 days), we reach 3 Chicchan 13 Pax, year 3 
Ben, the terminal date of the black series. 

Neither of the series in the right-hand serpent of this plate brings 
the count to either of the dates which stand below them. As yet I 
am unable to find in what the error consists. As the text above this 
right-hand serpent has been obliterated, it is possible, though I do not 
think probable, that a different initial date is given. As both series 
counted backward reach a 9 Kan, but of different months, I am 
inclined to believe that the error consists in one or both month sym- 
bols of the terminal dates. 

The other series which has not been considered is the red one in the 
serpent on plate lxix, Dresden codex. The difficulty in this case 
arises from the insufficient data, the number in the katun place 
having been omitted or obliterated, and the month symbol of the 
terminal date being too nearly obliterated for anything positive in 
regard to it to be determined. If the month symbol could be deter- 
mined by inspection, the data would be sufficient to give the num- 
ber of katuns correctly; but with the series in its imperfect con- 
dition, we can only state that, by a trial substitution of the numbers 
from 1 to 19 in the katun place, we find that this number must be 
either 1 or 14. If it be 1, the terminal date is 9 Ix 12 Zip; if it be 14, 
the terminal date is 9 Ix 12 Zac. As the fragment of the month 
symbol, small as it is, apparently forbids the supposition that it is 
Zac, it is probably Zip. 

Taking the difference between the red series, as thus corrected, and 
the black series, and counting back from 9 Ix 12 Zip, we reach 4 Eb 
5 Chen, year 9 Lamat, which agrees with the result of counting the 
black series from the initial date. The solution, therefore, appears to 
be satisfactory. 

As Dr Seler raises another question in regard to these high series 
of the Dresden codex, we will consider it before passing on. It will 
be noticed that in the text (double column) at the left of the ser- 
pents on plalcs lxi and lxix, there is, in each case, a numeral series 
given in symbols in the same form as those in the inscriptions. The 
one on plate lxi is 15-9-1-3, or L5 katuns, 9 ahaus, 1 chuen, 3 days; 
that on plate LXIX is 15-9-4-4, or L5 kal nns, !) ahaus, 4 chuens, I days. 
The date following in each case is 1) Kan 12 Kayab, and the date 
preceding is in each case is I Ahan 8 Cumhu. Now, Dr Seler, if I 

246 MAYAN CALENDAR SYSTEMS [eth. ann. 22 

rightly understand him, contends that this series belongs to, or is con- 
nected with, the series in the serpent figures, and is to show that the 
count is carried back to 4 Ahau 8 Cu nihil as the initial date, though 
he has failed to make connection between the dates by the series in 
the text. 

As the initial and terminal dates (4 Ahau 8 Cumhu and 9 Kan 12 
Ka} T ab) are the same on both plates, and the number of the katuns 
and ahaus the same in both, it is certain there is a mistake in one or 
the other in regard to the number of chuens and days — one being 4 
chuens, 4 days, and the other 1 chuen, 3 days — as the terminal date 
can not occur twice in the lapse of time between one and the. other, 
that is, in 61 days. However, neither series will connect the two dates. 
The series on plate lxix when reduced to days is as follows : 


15katuns ' -__ 108,000 

9 ahaus 3, 240 

4 chuens - . _ . 80 

4 days . _ _. 4 

Total '_ r 111, 324 

Subtract 5 calendar rounds 94, 900 

Remainder 16, 424 

Counting this number of days forward from 4 Ahau 8 Cumhu, year 
8 Ben, we reach 9 Kan 7 Cumhu, year 1 Ezanab — a date 37 years 
later than the proper one; nor will counting backward give the 
proper result. It is apparent from the problem itself that the 
numeral series must be materially changed in order to connect these 
dates, if this was the object of the aboriginal artist. That the two 
da'tes are too prominent for either to be changed will be admitted. 
As 4 Ahau 8 Cumhu falls in the year 8 Ben, and 9 Kan 12 Kayab in 
the year 3 Ben, the lapse of time from the former to the latter, count- 
ing forward (the necessary direction on Seler's assumption) is 2,904 
days (plus any number of calendar rounds); while the number of 
daj^s over and above the calendar rounds in one of the series (plate 
lxix) is 16,424 days, and in the other (plate lxi) is 16,263 days. The 
difference between 16,263 and 2,904 is 13,459. Therefore, correcting 
the series, as the dates can not be changed, involves dropping out 
13,459 days, or nearly 37 years. It is impossible to make this cor- 
rection by any change in the number of chuens and days, and as 
the katuns and ahaus are the same on both plates, it is presum- 
able that they are as they were intended to be. Therefore, while the 
positions of the dates in the text in relation to the numeral series 
would seem to indicate that they were intended to be connected by 
it, no justifiable correction or reasonable manipulation of the series 
appears to bear out this theory. It would seem from these facts that 
the data do not sustain Seler's assumption. 


Suppose, however, that it was the intention of the aboriginal artist 
to connect the dates by these short series, and that each of them con- 
tains some error, and when corrected would make the connection, let 
us see what the result would be. The entire series on plate lxix — 
taking that in the text as it stands, and the black one in the serpent 
figure, making 15-9-4-4 plus 4-5-19-13-12-8, 4 Eb ? (month)— would 
throw back the initial date 12,493,052 days, or a little over 34,226 
years, previous to the terminal date 4 Eb of the series. This is 
wholly inconsistent with the idea expressed by Seler (quoted farther 
on) that the terminal dates of the inscriptions indicate, respectively, 
the time of the erection of the monument, and that these dates fall 
within or after the 10th cycle (Goodman's 9th of the 54th great cycle). 
If the 4 Ahau 8 Cumhu of this series is the same 4 Ahau 8 Cumhu in 
actual time as the first of Goodman's 54th great cycle — or, as Seler 
calls it, the "normal date" — then the series must run far into the 
actual future, or all the dates of the inscriptions must be far back in 
the past, and are merely theoretical. The only other supposition is 
that the 9 Kan 12 Kayab in the columns at the left is not identical 
with the 9 Kan 12 Kayab that stands above the serpent, and with 
which the series in the folds are undoubtedly connected. 

As the final date in the series referred to in the preceding para- 
graph is incomplete, in lacking the day of the month, we will try the 
one on plate lxi. Using the black series in the folds of the left ser- 
pent, as this is the largest of the four great series on this plate and 
hence presumably the last (though the rule, if correct, should hold 
good with any of the series), we have 15-9-1-3 plus 4-6-14-13-15-1 (as 
they stand on the plate). Counting 20 cycles to the great cycle and 
changing to days, we arrive at the following result: 


4 great cycles 11,520,000 

6cycles 864,000 

14 katuns 100, 800 

13 ahaus 4, 680 

15 chuens 300 

1 day 1 

12, 489, 781 
Add amount of short series 111 , 263 

Total 12,601,044 

Subtract 663 calendar rounds 1 2 , 583 , 740 

Remainder 17,304 

Using this remainder and counting forward from 4 Ahau 8 Cumhu, 
year 8 Ben, we reach 9 Imix 9 Mol, year 4 Ben. This is wrong. Let us 
use the series as corrected on a previous page, to wit: 4-6-14-11-3-1, 

248 MAYAN CALENDAR SYSTEMS [eth. ann. 22 

or 12,488,821 clays. Adding the shorter series and counting forward 
from 1 Ahau 8 Cumhu, we reach the date 4 Kan 2 Yaxkin. This 
again is wrong. Using the larger series as corrected and counting 
from 9 Kan 12 Kayab we reach, as has already been shown, the cor- 
rect date, 3 Chicchan 13 Pax. It is therefore fair to conclude that 
there are no sufficient grounds for Seler's supposition. 

These erroneous conclusions arise chiefly from the mistaken idea 
that these numbers, ahaus, katuns, etc., are real time periods. More- 
over, it does not necessarily follow, where such high numbers are 
used, that 4 Ahau 8 Cumhu is what Seler calls the "normal date"; 
that is to say, the initial day of Goodman's 54th great cycle. But this 
does not matter in the present case, as the date can not be connected 
with any of the others given in the series. 

Even could the series be reasonably changed so as to make the con- 
nection between the given dates, we still have staring us in the face the 
fact that 9 Kan 12 Kayab is actually and beyond question used in the 
codex as the initial day of the so-called great cj^cle in six instances, 
and that a Kan is the initial date in 3 times 6 other instances. It is 
true that these so-called great cycles are but orders of units, steps in 
numeration, and not real time periods; nevertheless, they are just as 
real when counting from a Kan as from an Ahau. 

In order that the reader may clearly understand the object in view 
in introducing these calculations, and see the bearing they have on 
the question, it is necessary again to refer to the basis of Goodman's 
theory of the Mayan time system, and especially of his supposed 
separate "chronological calendar." 

Goodman maintains that in addition to their regular annual cal- 
endar in which time was counted by years, months, days, etc., the 
Mayas made use of another time system which he terms the "chrono- 
logical calendar." In this system, according to his theory, they 
counted time by certain determinate periods, which, according to the 
nomenclature arbitrarily adopted by him, are termed chuens (each of 
20 days); ahaus (each of 18 chuens or 360 days); katuns (each of 20 
ahaus or 7,200 days); cycles (each of 20 katuns or 144,000 days); 
great cycles (each of 13 cycles or 1,872,000 days), and a grand era 
equal to 73 great cycles. These he believes to be real time periods, 
as truly so as the years, etc. , of the annual calendar, systematically 
arranged and all above the chuens always (so far as time count in 
the inscriptions is concerned) beginning with a day Ahau, the great 
cycles always with the day 4 Ahau. It is in this supposition that 
Goodman's great error lies, and, in order to support his premise, he 
changes two of the steps of the Mayan numeral system without the 
slightest evidence on which to base the change, and he also introduces 
factors into the numeral system which are wholly unknown to it. If 
these statements which I make can be maintained by satisfactory 
evidence, then his theoretic "Archaic Chronological System" falls to 

thomas] Goodman's "archaic chronological calendar 77 249 

the ground, though his discoveries as to the signification of certain 
glyphs and the manner in which they were used be genuine, and his 
calculations of series be correct, and though his tables be also correct 
in the main. 

The annual calendar system, which is that one long ago explained 
and accepted (that of months, years, etc.), is not in dispute. It is 
his theory of another time system, his so-called "Chronological Cal- 
endar," which I assert is without basis of fact. This calendar, which 
he says he "finally deduced," he expects will be challenged, but he 
"leaves it to defend itself, conscious that it is as infallible as the 
multiplication table." 

Before referring to the proof bearing on this subject already pre- 
sented, we shall call attention again to Goodman's method of num- 
bering these periods. The chuens he says were numbered 18, 1, 2, 3, 
etc., up to 17; the ahaus and katuns were numbered 20, 1, 2, 3, etc., 
up to 19; the cycles, 13, 1, 2, 3, etc., up to 12; and the great cycles, 
73, 1, 2, 3, etc., up to 72. On this subject he remarks as follows: 

Another consideration which must be constantly borne in mind is that all Maya 
dates relate to elapsed time. When a date is given it must be remembered that 
it is not the beginning of a period yet to run its course, but the beginning of one 
denoting a period already concluded. The ingenious numeration of their periods 
was designed to prevent confusion in this regard. The first day, chuen, ahau, 
katun, cycle, and great cycle is not numerated 1, but 20, 18, 20, 20, 13, 73, as the 
case may be, denoting that the full round of the period has run and that this is 
the commencement of a new count. In other words, these beginning numerals 
are equivalent to naught or no count, the periods being designated only until after 
they had fully passed. It is very difficult to keep track of this style of numera- 
tion — so difficult, in fact, that familiar as I am with it I am distrustful of having 
made some lapses in these pages. 

That he has made a mistake in this statement, in order to fit the 
facts with his theory, and that he carries this mistake throughout his 
entire work, is easily shown, and will appear from what follows. 

That the count is forward to some date in the future, as compared 
with the initial date, in most of the series of the inscriptions, is appar- 
ent from the examples given by Mr Goodman in his work; and that 
it is forward to some future date, as compared with the initial day, in 
eveiy initial series, must be admitted. Therefore, his assertion can 
not be intended to contradict this fact. What he intends to declare 
is this, that when a date is given, as the first day of the 2nd katun or 
ahau, we must understand that it is really the first day of the 3rd 
katun or ahau, the 2nd being completed; or when 2 ahaus and 3 
chuens are mentioned, we are to understand 2 completed ahaus and 
3 completed chuens. 

Let us see if we can ascertain how this strange method of number- 
ing these so-called periods originated. It must be remembered that 
this numbering is the consecutive numbering, as that of the days of 

250 MAYAN CALENDAR SYSTEMS [eth. ann. 22 

the month, and not the numbering (in the 13 series) of the day Ahau 
as mentioned above. I quote again from his work (pages 12 and 13): 

Poor Don Pio! To have the pearl in his grasp and be unaware of its priceless- 
ness — like so many others! But I must not exult too much yet. The succession 
of the katuns, reckoned according to this principle, is yet to be ascertained before 
my fancied discovery can be established by a crucial test. I score the ahaus off 
in the foregoing order, and, sure enough, the twentieths give the desired result: 
11. 9, 7, 5, 3, 1, 12, 10, 8, 6, 4, 2, 13. Eureka! The perturbed spirit of the Maya 
calendar, which has endeavoured so long to impart its message to the world, may 
rest at last. 

But, though confident I had discovered the secret of the ahau and katun count, 
when I tried the plan on the dates and reckonings of the inscriptions it proved 
totally inapplicable. There were periods into whose nature I had no insight, and 
if those I surmised to be ahaus and katuns were really so the former would not 
corns in the right order, while the latter were excessive and numerated in a way 
quite unintelligible. It was discouraging, but I did not lose faith in my discov- 
ery. The inapplicability of the Yucatec scheme to the reckonings of the inscrip- 
tions, probably, was simply owing to different methods of computing the ahaus 
and katuns. There was no alternative but a patient and exhaustive analysis of 
the Archaic dates and time reckonings. 

It would be tedious as useless to recount trials — failure outranking success a 
thousand fold — the results of which constitute the bulk of this book. I will only 
state, in brief, that I determined the character of the chuen and great cycle 
periods; that I discovered the first chuen was numerated 18, the first ahau, katun, 
day and day of the month, 20, and that the first cycle of the great cycle was num- 
bered 13 — the unit attaching to the second period in all instances; that I ascer- 
tained the cycle was composed of twenty katuns, numerated 20, 1, 2, 3, etc., up to 
19, instead of according to the Yucatec order; that I finally deduced a chronolog- 
ical calendar whose perfect accord with the principal dates and reckonings 
throughout the inscriptions is proof of its correctness, and by reversing the process 
succeeded in reconstructing the outlines of the entire Archaic chronological 
scheme. I expect my calendar to be challenged. It would be without precedent 
in the history of discovery if it were not. But I leave it to defend itself, con- 
scious that it is as infallible as the multiplication table, and knowing that all 
antagonists must finally go down before it. 

By reading between the lines of this quotation, and noting the dif- 
ficulties he encountered, we readily see that his theory was outlined 
before the difficulties presented themselves. Why should he find it 
necessaiy to number the first chuen 18, the first ahau 20, and the first 
cycle 13 were this not so? Take the short series 13-9-9 from 12 
Caban 5 Kayab to 6 Cimi 4 Tzec, which he mentions, and says Vvorks 
out all right. There is no difficulty if we count it 13 ahaus plus 9 
chuens plus 9 days, just as we might say 13 hundreds 9 tens and 9. 
If we read it as it really is, 13 units of the 3rd order (360 each) plus 
9 units of the 2nd order (20 each) and 9 units of the 1st order (1 each), 
there is no difficulty in showing that it is an exact measure of the 
lapse of time between the given dates. 

The difficulty, as we may safely assume, arose from the fact that 
the count would not fit in with the theory he had formulated but had 

thomas] Goodman's archaic calendars 251 

not perfected. He had probably outlined the tables of his "Archaic 
Chronological Calendar," but instead of numbering them as we find 
them now given in his work, the cycles were numbered 1, 2, 3, etc., up 
to 13; the katuns, 1, 2, 3, etc., to 20, etc. Conceiving the idea that 
the numbers in the series (as the 13-9-9) should express the numbers 
in his scheme — that is to say, should be read the 13th ahau, the 9th 
chuen, and 9th day — he found that it would not give the correct result. 
Here indeed was a difficulty, a difficulty of fitting facts to a theory, but 
not one in reality, for the series taken as it stands works out correctly. 
In order to overcome this difficulty and at the same time save his 
theoiy he seemingly hit upon the ingenious device of a supposed 
Mayan method of numbering periods somewhat as the surveyor num- 
bers his stations, beginning with (naught), or what gives the same 
result and avoids the use of the cypher, which he contends was not 
used by the Mayas, of bringing forward the last number of the 
preceding period to be the first of the one following. Thus in his 
"Archaic Annual Calendar" he has pushed down one step the true 
dominical days, Akbal, Lamat, Ben, Ezanab, although retaining 
their proper numbers, and has brought forward, with the number 20 
attached, the preceding days, Ik, Manik, Eb, Caban, and begins the 
numbering of the chuens with 18, of the ahaus and katuns with 20, 
etc. This, of course, overcomes the difficulty, as what is numbered 
the first ahau, etc., is, in fact, the second, and in the example given 
the 13th ahau is, in fact, the 14th, and the 9th chuen the 10th, and 
hence, by his method of numbering, the 13th ahau, 9th chuen, 9th 
day is equivalent to 13 complete ahaus, plus 9 complete chuens, plus 
9 days. This plan will undoubtedly preserve the proper order of 
succession. The only real errors it introduces, if considered merely a 
method of numbering, is in making the wrong days dominicals and 
in carding the last day of one month forward to become the first 
day of the next, one or two examples of which are pointed out in my 
previous paper. These examples have since been more fully dis- 
cussed by Mr. Bowditch, witli the result of strongly inclining him to 
accept Goodman's theory in this respect. They are noted in my 
Maya Year (figure 20), though not discussed there as to tiie point 
here raised. 

As further evidence bearing on this question, I add the following: 
There is no such method of numbering found in the inscriptions, or 
in the codices, Mayan or Mexican, unless in the examples above 
referred to, and there is no such method mentioned by any of the 
early writers. Perhaps, however, the most important point to be 
decided in this connection is the query, Did the Mayas in fact 
number these so-called periods? How many were to be taken was 
indicated by symbols, but there is no evidence, so far as I am 
aware, that they were numbered, except in a single instance found 

252 MAYAN CALENDAR SYSTEMS [eth. ann. 22 

on the north and south faces of Stela J at Copan. Here, it is true, 
we find a succession of ahau symbols of the usual type, placed in 
somewhat regular order and numbered in regular succession from 1 
to 16, beyond which the remaining glyphs (only two, however) are 
obliterated. Whether these numerals are intended as a successive 
numbering or intended merely to indicate so many ahaus, is not 
known ; however, it looks like regular numbering, and is so accepted. 
But, unfortunately for Goodman's theory, the series clearly begins 
with number 1. To get around this difficulty he assumes that it is 
to be understood that 1 ahau has passed, yet he admits that the 
symbol on that numbered 1 signifies "beginning." Thus the only 
example of numbering these so-called periods found in all the records 
is emphatically against his theory, in order to sustain which he 
literally begs the question by saying it must be assumed as under- 
stood that 1 ahau has passed. We are justified, therefore, in regard- 
ing his scheme of numbering as wholly unnecessary to explain the 
numeral and time series of the inscriptions, for considering his 
so-called time periods merely orders of units will give a full explana- 
tion, so far as the counting is concerned, in every case. 

But these items do not show all the errors in the above-quoted 
statement from Goodman's work. That but 13 cycles were counted 
to the great cycle, I have shown by mathematical demonstration is 
untrue, so far, at least, as the Dresden codex is concerned. I have 
shown that this codex, instead of counting 13 cycles to the great cycle, 
counts 20, thus following regularly, as would naturally be supposed, 
the vigesimal system. It is true that Goodman admits that the codices 
belonging to what he calls the Yucatec group not only count 20 cycles 
to the great cycle, but count from some three or four different initial 
days. This admission, however, does not avail him anything in the 
way of clearing his theory of the difficulty presented. In the first 
place, the Dresden codex can not be classed with the so-called Yucatec 
group. This group, which includes the Troano and Cortesian codices, 
and the codex used by Landa, makes Kan, Muluc, Ix, and Cauac the 
dominical days; while the Dresden codex, from which the examples 
given above showing the use of 20 cycles to the great cycle were 
taken, follows the system of the inscriptions in using throughout 
Akbal, Lamat, Ben, and Ezanab as dominical days. Moreover, it 
gives high series wholly unknown to the Troano and Cortesian codices; 
and it introduces in some three or four places, as numerical charac- 
ters, precisely the same symbols as those of the inscriptions named by 
Goodman katun, ahau, and chuen, and in one or two places uses a 
face character to represent the ahau. 

What grounds, therefore, can Goodman have for asserting that the 
system used in the inscriptions is different from that used in the Dres- 
den codex, which he evidently includes under the term "Yucatec 




system " ? There is nothing in either the Troano or Cortesian codex 
by which to determine the number of cycles they count to the great 
cycle. What sj'stern was used in the Yucatan inscriptions is not posi- 
tively known, but, as is shown below, thej T probably agreed with the 
Troano and Cortesian codices. Goodman says he has been unable to 
find a single Yucatec inscribed date. After careful inquiry and 
examination of the casts of inscrip- 
tions in the chief eastern museums 
and all the photographs, drawings, 
and figures in reach, without finding 
one, I have had my attention called 
by Mr Saville, of the New York 
Museum of Natural History, to a 
photograph by Mahler, taken at Xca- 
lumkin, in Yucatan, which is repro- 
duced in Le Plongeon's "Queen 
Moo," which, if I correctly interpret 
it, may be an indication of the sys- 
tem used in the Yucatec inscrip- 
tions. This is shown in figure 157 
from a copy of the photograph kindly 
furnished by Mr. Saville. 

The day (All) is evidently 8 Caban, 
the 4th day, apparently, of the month 
Zotz, though the month symbol is 
somewhat unusual in form. If the 
day symbol is properly interpreted 
Caban 5 of which there can scarcely 
be a doubt, then, as the -1 dots over 
the month symbol are very distinct, 
it is certain (whether we can deter- 
mine the month symbol or not) that 
the year must begin with the day 
Ix, hence the dominical days must 
be Kan, Muluc, Ix, and Cauac. 
This is the calendar svstem of the 
Troano and Cortesian codices and 
also of the codex followed by Landa. 
This result I must confess is con- 
trary to m}^ expectation and carries 
back the Yucatec calendar system to the days of the inscriptions. It 
is true thai a single inscribed date is a slender basis on which to 
reach a decision, but we must accept it until other evidence on the 
point is forthcoming. Goodman suggests that the Cocomes, Xius, 
Chels, and Itzas had eacli their own "chronological system, using a 

Fig. 157. Inscription at Xcalumkin, Yu- 
catan. From a photograph by Mahler. 



[ETH. ANN. 22 

common calendar." On what he bases this opinion, which is equiva- 
lent to saying they had different numeral systems, I am not aware. 
That the system in vogue at Tikal (in the Itza region of the Peten 
district) was the same as that of the inscriptions at Palenque, Copan, 
and Quirigua is well known. 

Let us return to the exceptional series of the Copan inscriptions 

mentioned above (west side of Stela N). Although 
it was discussed at some length in my previous 
paper, a reexamination has brought to light some 
facts overlooked in the first examination, which 
have an important bearing on the question in- 
volved; and they will be noticed here. This series 
reversed is as follows: 14-17-19-10-0-0 to 1 Ahau 8 
Chen (figure 158). Written out it is 14 great 
cycles, 17 cycles, 19 katuns, 10 ahaus, chuens, 
days, to 1 Ahau 8 Chen. Changed into days it 
gives the following result, counting 20 cycles to 
the great cycle : 


14 great cycles I 40, 320, 000 

17 cycles 2,448,000 

19katuns____ 136,800 

10 ahaus : 3, 600 


Total 42,908,400 

Subtract 2,260 calendar rounds 42, 894, 800 


13, 600 

If we count back this number of days from 1 
Ahau 8 Chen, year 3 Ben, it brings us to 12 Ahau 
13 Zotz, year 5 Lamat, which will be the first day 
of the first, or most remote, of the 14 great cycles, 
counting the series in this manner upward from 
the loth: 

Fig. 158. Part of in- 
scription on the west 
side of Stela N, Co- 
pan. Maudslay, part 
4, plate Lxxix. 

1st great cycle 

2nd great cycle, etc. 

14th great cycle 

(15th great cycle) 

17 cycles 

19 katuns 

10 ahaus 




If we count back from the same date (1 Ahau 8 Chen) the 17 cycles, 
19 katuns, and 10 ahaus, we reach the first day of the (incomplete) 
15th great cycle as we have numbered them above. This day is 5 


Ahau 8 Cumhu, year 9 Ben. If we count back the great cycles one 
by one (counting 20 cycles to a great cycle), we shall find the initial 
dates to be as follows — the numbers given the great cycles being, of 
course, arbitrary: 

1st great cycle 12 Ahau 13 Zotz, year 5 Lamat 

2nd great cycle 5 Ahau 3 Ceh, year 4 Ezanab 

3rd great cycle 11 Ahau 8 Pop, year 4 Ben 

4th great cycle 4 Ahau 18 Mol, year 3 Akbal 

5th great cycle 10 Ahau 8 Pax, year 2 Ben 

6th great cycle 3 Ahau 13 Tzec, year 2 Lamat 

7th great cycle 9 Ahau 3 Mac, year 1 Ezanab 

8th great cycle 2 Ahau 8 Uo, year 1 Ben 

9th great cycle 8 Ahau 18 Chen, year 13 Akbal 

10th great cycle 1 Ahau 8 Kayab, year 12 Ben 

11th great cycle 7 Ahau 13 Xul, year 12 Lamat 

12th great cycle 13 Ahau 3 Kankin, year 11 Ezanab 

13th great cycle 6 Ahau 8 Zip, year 11 Ben 

14th great cycle 12 Ahau 18 Yax, year 10 Akbal 

15th great cycle 5 Ahau 8 Cumhu, year 9 Ben 

This result shows our calculation to be correct, taking the day of 
the inscription (1 Ahau 8 Chen) as that from which to count back. 
As there are 14 complete great cycles, which we estimate at 20 cycles 
each, and the minor periods (17 cycles, 19 katuns, and 10 ahaus), the 
latter must fall in the 15th great cycle, which is incomplete. Count- 
ing back these minor periods, we reach, as has been stated, 5 Ahau 8 
Cumhu, year 9 Ben, as the first day of this 15th great cycle. Counting 
back from this latter date 20 cycles (or 1 great cycle) we reach 12 
Ahau 18 Yax, year 10 Akbal, the first day of the 14th great cycle, 
and so on to the initial day of the first, which we find to be 12 Ahau 
13 Zotz, year 5 Lamat, giving exactly the same result as our calcula- 
tion of the whole as one single series. By both methods the first day 
of the entire series, and hence the first great cycle as numbered 
above, is found to be 12 Ahau 13 Zotz. But this, though correct so 
far as calculation is concerned, is not proof, as the results given must 
necessarily follow if the date counted from is 1 Ahau 8 Chen, and 2Q 
cycles are counted to a great cycle. This is unsatisfactory, as it fails 
to bring in as the first day of a great cycle 4 Ahau 8 Cumhu, which 
was a normal date at Copan. 

I am strongly inclined to believe that the terminal date of the series 
instead of 1 Ahau 8 Chen, as given in the inscription, should be 13 
Ahau 8 Chen, which falls in the year 2 Ben. If we count back from 
this date 17 cycles, 19 katuns, 10 ahaus, chuens, days, it will bring 
us to 4 Ahau 8 Cumhu, year 8 Ben, as the first day of the 15th great 
cycle, as we have arbitrarily numbered them above. If we count back 
the entire series, 14-17-19-10-0-0, from 13 Ahau 8 Chen, year 2 Ben, it 
brings us to 11 Ahau 13 Zotz, year 4 Lamat, as the first day of the 1st 

256 MAYAN CALEKDAK SYSTEMS [eth. ann. 22 

great cycle as numbered above. The first days of the great cycles 
would then be as follows: 

1st great cycle 11 Ahau 13 Zotz, year 4 Lamat 

2nd great cycle 4 Ahau 3 Ceh, year 3 Ezanab 

3rd great cycle 10 Ahau 8 Pop, year 3 Ben 

4th great cycle 3 Ahau 18 Mol, year 2 Akbal 

5th great cycle 9 Ahau 8 Pax, year 1 Ben 

6th great cycle 2 Ahau 13 Tzec, year 1 Lamat 

7th great cycle 8 Ahau 3 Mac, year 13 Ezanab 

8th great cycle 1 Ahau 8 Uo, year 13 Ben 

9th great cycle 7 Ahau 18 Chen, year 12 Akbal 

10th great cycle 13 Ahau 8 Kayab, year 11 Ben 

11th great cycle 6 Ahau 13 Xul, year 11 Lamat 

12th great cycle 12 Ahau 3 Kankin, year 10 Ezanab 

13th great cycle 5 Ahau 8 Zip, year 10 Ben 

14th great cycle 11 Ahau 18 Yax, year 9 Akbal 

15th great cycle 4 Ahau 8 Cumhu, year 8 Ben 

The method of numbering the great cycles must be understood as 
wholly arbitrary, given merely for convenience, and to include the 15 
that are referred to in the count. I do not believe that there was any 
consecutive numbering of these supposed time periods in the sense 
indicated by Goodman; in fact, as I expect to show, they were not 
time periods in any true sense of the term. 

The reason for believing that the date following the inscription 
should be 13 Ahau 8 Chen instead of 1 Ahau 8 Chen is that 4 Ahau 8 
Cumhu, as appears from the inscriptions at Copan and Quirigua, 
was the favorite initial date, most of the initial series going back 
to it, and that counting back the minor periods of the series from 
13 Ahau 8 Chen brings us to 4 Ahau 8 Cumhu. If we turn to 
Goodman's "Archaic Chronological Calendar' 1 and count forward, 
from the beginning of his 54th great cycle, 17 cycles, it will bring us 
to the 4th cycle of his 55th great cycle, and to the 19th katun of this 
cycle and the 10th ahau of this katun, where we find the day to be 
13 Ahau 8 Chen. We are therefore of the opinion that the terminal 
day of the long series should be 13 Ahau 8 Chen, and that Goodman 
is wrong in rejecting it. As there are 17 cycles, it proves, as it stands, 
that the authors of the inscriptions counted 20 cycles to the great 
cycle, which is consistent with their system of numeration. I have 
shown in my previous paper why 1 Ahau 8 Zip can not be the initial 
date of this series. 

As bearing on the explanation of this series, the following facts in 
regard to the symbols are worthy of special notice. It will be seen 
by an inspection of the series shown in figure 158 that the great C} T cle 
symbol (glyph 5) is a face character very much like that of the cycle, 
except that it has a superfix, which unfortunately is too nearly oblit- 
erated to be traced. However, it is noticeable that in both it and the 
cycle symbol the hand figure is across the lower jaw. According to 
Goodman, "the hand on the cheek, the thumb or wrist forming the 


lower jaw, usually characterizes the face sign for 20 " (page 52), and 
this conclusion is sustained by the evidence we have given above. 
Goodman's perverseness in contradicting his own evidence in order to 
maintain his theory is shown in reference to this sign. It is found 
almost universally on the cycle face characters, as may be seen in his 
examples on page 25 of his work. It is true that it may be contended, 
as Goodman in fact does contend, that it signifies that 20 of the next 
lower order make one of this order. Admit this; it follows that when 
the same sign is found on the great cycle symbol, it signifies that 20 
of the next lower order (or cycle) make one great cycle. Although 
but one example of the great cycle face symbol has been found, it 
bears clearly and unmistakably this hand sign, and not only is this 
not denied by Goodman, but is accepted by him and copied as an 
example of the symbol of this period on page 25 of his work. 

Thus it will be seen that from whatever side we view the evidence 
bearing on this question, it is against . Goodman's theor} T of only 13 
cycles to the great cycle. However, before closing the discussion of 
this point I desire to call attention to one other series, found on Stela 
of Quirigua, which seems to have a bearing on the question. This 
is as follows: 54-13-0-0-0-0, 4 Ahau 8 Cumhu — in other words, 54th 
great cycle, 13 cycles, katuns, ahaus, chuens, days, to 4 Ahau 
8 Cumhu, the 13 being the ordinary numeral symbols, dots and short 
lines, and very distinct. Goodman's only comment (page 127) is, 
"This date is the beginning of the 54th great cycle." As 4 Ahau 
8 Cumhu, is, according to his reckoning, the initial day of the 54th 
great cycle, the series, according to this explanation, covers no lapse 
of time whatever. Yet, according to his theory, the numbers in these 
series always relate to time which has elapsed. Hence the 13 cycles 
relate to 13 of these so-called periods which have passed and still 
signify no time whatever. This is a palpable contradiction into 
which he has been led in his effort to maintain an erroneous theory. 
If he had written the series " 53-13-0-0-0-0 to 4 Ahau 8 Cumhu, 1 ' it 
would have been correct so far as the count is concerned. 

Dr Seler in his able article, " Die Monumente von Copan und Qui- 
rigua und die Alter-Platten von Palenque " (Zeitschrif t fur Ethnologie, 
Heft 6, 1899, pages 670-738), makes some remarks in regard to the 
series above noticed to which it is desirable to call attention. a 

It appears from this article that he follows Goodman in counting 
13 cycles to the great cycle, or 13 units of the 5th order to make one 
of the sixth (I repeat again that Goodman's terms are used merely 
for convenience). Moreover, he seems to look upon these as real 
time periods. That he, who is so familiar with the subject, has not 

a This article was not received by me until all this paper, except the last few pages and the 
notices of it which have been inserted, had been written. As I have seen no reason, because of 
Seler's article, to change anything previously written, I make this statement as due to myself. 

22 eth— 04 17 

258 MAYAN CALENDAE SYSTEMS [eth. ann. 22 

entered into a careful examination of the basis on which Goodman's 
theoretical ' ' Chronological System " rests, and that he has accepted 
Goodman's theory of 13 cycles to the next higher period, without 
thoroughly testing it, and noting the 20 cycles of the Dresden codex, 
is somewhat surprising to me. However, he may have reserved the 
discussion of these points for a future article. 

In speaking of the series last referred to, 54-13-0-0-0-0, 4 Ahau 8 
Cuinhu, he says: 

Here one sees that the final date is the normal date itself. Its distance from 
the normal date can be placed only at or the above-named immense period of 
18,720 years. The builders of the monuments have done neither. They have 
provided all the lower multiplicands, or smaller periods, with the index 0, but to 
the highest and greatest they have placed the multiplier 13. Thirteen is the 
number of the index figures which are possible with the tun, the katun, and the 
cycle names. If, consequently, here at the beginning of the initial series the 
thirteen cycles are named, nothing else is meant than "the periods or epochs 
generally." And the whole initial series would consequently give about the fol- 
lowing idea: " This is a chronological monument. The beginning of the number- 
ing is the day 4 Ahau 8 Cumku." And the fact that on the west side of the 
same stela another, definite date and its distance from the normal date is named 
agrees very well with this. 

Similarly, in my opinion, are to be understood the thirteen 'cycles which are 
chiseled on the two sides of Stela C of Copan, immediately under the katun signs, 
the initial and chief hieroglyphs. 

It seems clear from this that he has adopted Goodman's interpreta- 
tion of the series, unaware of its incongruity with the interpretation 
of all the other initial series, and the fact that it stands in opposition 
to his own conclusion stated a little farther on in the same article. 
As proof of the latter statement, I refer to the quotation from his 
article given hereafter (page 292). 

JN~ow, it is apparent that, if the series be interpreted as signifying 
no lapse of time, but as a mere assertion that the date of the event 
commemorated was 4 Ahau 8 Cumhu the first day of the 54th great 
cycle, which interpretation Seler adopts, then the monument must 
have been erected 3,550 years before the beginning of the cycle 
which he numbers the tenth (Goodman's ninth). It is apparent, there- 
fore, that he has failed to see the contradiction between this state- 
ment and that which places the erection of the monuments of Copan 
and Quirigua in the tenth cycle. He objects to the lapse of 3,160 
years between the erection of the monuments of Palenque and those 
at Copan and Quirigua, as improbable, but here admits, by his inter- 
pretation, a lapse of 3,550 years between monuments at Quirigua. 

I have stated above that Goodman's so-called time periods, chuens, 
ahaus, katuns, etc., are in reality nothing more than orders of units, 
or steps in numeration. Although this point has been discussed to 
some extent in my previous paper, I will add here some further 
evidence bearing on it. 




As a means of illustrating the use of numbers by the Mayas, in 
relation to time, the following example — which is part of a series on 
plate Lix of the Dresden codex (figure 159) — is presented: 

• • • 

13 Cahan 

13 Canac 

13 Imix 

• • • 

13 Akhal 

13 Chicchan 

13 Manik 

As this series ascends toward the left hand the forward count will 
be in that direction. Starting with the column at the right hand, we 
subtract it (3-18) from the next one to the left, and this one from that 
immediately to the left of it, and so on to the last. 

The difference in each case is found to be 3-18; that is, 3 twenties 
(3x20) plus 18 equal 78 days, the day being the unit. Counting for- 

wm mumm wmMMH w MZ-f 

Fig. 159. Lower division of plate lix, Dresden codex. 

ward 78 days from 13 Manik of any year (say 13 Manik 20 Zotz, j ear 
12 Lainat) we reach 13 Chicchan (in this case 18 Mol, same year). 
Counting forward 78 days from the last date we reach 13 Akbal Hi 
Ceh, same year; 78 more (always counting from the last date), 13 
Imix 11 Pax, same year; 78 more, 13 Cauac 7 Uo, year 13 I Jen. If we 
count back 78 days from 13 Manik 20 Zotz (first column at the right 
hand), we reach 13 Muluc 2 Pop, year 12 Akbal, which is the initial 
day of the whole series, the month and year of the first given day 
being as assumed above. 

Attention is called to this series not because it presents any peculiar 
feature, but to show that considering the numerals merely as num- 
bers in respective orders of units will furnish a full and satisfactory 
explanation of their objed and use. I take for granted that the 
simplest explanation, if it meets every requirement and presents 
nothing inconsistent with the known facts regarding the Maya time 

260 MAYAN CALENDAR SYSTEMS [eth. ann. 22 

and numeral systems, should be accepted rather than a theoiy which 
introduces new and hitherto unknown features. 

If we use ordinary numbers in place of the numeral symbols, and 
keep them in the relative positions given above, the result will be as 
follows ; 

3rd order of units. .. 


2nd order of units 






1st order of units 






13 Cauac 

13 Imix 

13 Akbal 

13 Chicchan 

13 Manik 

If we assume these to be successive orders of units indicated by 
relative position, increasing upward, the day being counted as the 
primary unit and the vigesimal system being used, except in passing 
from the second to the third order, where the multiple is 18, all the 
requirements of the series will be met. Thus, in the first, or right- 
hand, column, we have 18 units of the 1st order and 3 of the second, 
or 3x20=60, making together 60+18=78; and in the second column 
16 of the 1st order and 7 of the 2nd order, or 7x20=140, making 
together 140+16=156, and so on to the fifth column, where we find 
10 of the 1st order, 1 of the second, or 1x20=20, and 1 of the 3rd 
order, or 18x 20=360, making together 360+ 20+ 10=390.' Thesenum- 
bers give correct results, as, counting 78 days from 13 Muluc, we reach 
13 Manik; counting 156 from the same initial date, we reach 13 Chic- 
chan, or counting 390 days, we reach 13 Cauac. It is clear, therefore, 
that if we take these numerals to indicate the number of different 
orders of units, the orders being indicated by relative position, all 
requirements of the series are satisfied and the proper results are 

If we take one of the high series we find that precisely the same rule 
obtains, as, for example, one of those on plate lxii, Dresden codex 
(see plate lxxix) : 

4 of the 6th order of units .... 4 

6 of the 5th order of units : 6 

9 of the 4th order of units - — : — : — i 9 

15 of the 3rd order of units ^^^^ 15 

12 of the 2nd order of units ' ' 12 

19 of the 1st order of units ' 19 

This is upon precisely the same principle as our method of expressing 
numbers, except that it is according to the vigesimal system, while 
we use the decimal. Take the number 643,527, where the relative 
positions express the relative values, it becomes possible to represent 
the number thus : 

6 hundred thousands 

4 ten thousands 
3 thousands 

5 hundreds 
2 tens 

7 units 


If, instead of adding- the written names, simply the figure should 
be given, the relative positions being maintained and understood, we 
would have the Maya method, and the value would be known as well 
as by our ordinary method of writing numbers horizontally. 

I have given these details of elementary rules and principles in 
order to lead up to this point, viz, that symbols may be used to indi- 
cate orders of units instead of position. In the last example given 
above, a symbol may be adopted for the "hundred thousands," 
another for "ten thousands," another for "thousands," etc. They 
may then be grouped in any regular order most convenient, and yet 
be as correct!}' read as by position. This is precisely what has been 
done in the inscriptions. Symbols have been adopted to indicate the 
orders of units, as it was inconvenient to do this by means of relative 
position alone with the dots and short lines — at any rate it is apparent 
that the latter method is not so well adapted to the glyph form in the 
inscriptions; but even here we see a strong tendency to maintain the 
relative position which almost universally obtains and is often the 
only means of determination. If we take Goodman's work and go 
through it from beginning to end and substitute in every series where 
they occur "units of the 2nd order" for his chuens, "units of the 3rd 
order" for his ahaus, "units of the 4th order" for his katuns, "units 
of the 5th order" for his cycles, and "units of the 6th order" for 
his great cycles, the result will be correct in every instance. I am 
fulh* aware that this will be true whether we call them real time peri- 
ods or orders of units. The point, however, for which I am contending 
is, that as the Mayas had a system of numeration and must have used 
it in expressing numbers in the codices and inscriptions, and this 
numeral system corresponds exactly with Goodman's supposed time 
periods so far as these are given numerically correct by him, there 
is no necessity or reason for the theory of a separate Maya chrono- 
logical system (identical so far as correctly given with the Maya num- 
eral system as used in counting time), differing from their calendar 

From the evidence given in the earlier part of this paper and what 
has been presented in my preceding paper, the following conclusions 
appear to be clearly justified: 

That Mr Goodman has discovered independently the signification 
and numeral values of the symbols found in the inscriptions which he 
designates by the names cycle, katun, ahau, chuen, and calendar 
round, though this had been already done in part by others. 

That he has discovered that certain face and other characters arc 
number symbols, and has ascertained their values. 

That he has determined the object and use of the numeral scries, 
and the method of counting by the same series from the preceding 
and following dales, as well as to them. 


It is also equally apparent that his theory of a Maya chronological 
system, distinct from the Maya calendar system — the Mayan method 
of numeration in counting time — and his method of counting 13 
so-called cycles only to the so-called great cycle and 73 great cycles 
to his so-called grand era are not justified by the facts, nor is his 
method of numbering the cycles, katuns, etc., beginning with 73, 13, 
and 20, satisfactorily proved ; and also that his selection of Ik, Manik, 
Eb, and Caban as the dominical days is erroneous, the true dominical 
days being Akbal, Lamat, Ben, and Ezanab, both in the inscriptions 
and Dresden codex. 

Let us turn next to his method of numbering the so-called great 
cycles. According to his theory, as we have seen, 73 great cycles are 
counted to what he calls the grand era, the common multiple of all 
the factors of the calendar system and supposed "chronological 
system." The reason why he adopted this theory is explained in my 
previous paper, and the explanation need not be repeated here, except 
so far as merely to state that in order to find a common multiple of 
the various time, periods, one must include the number 365, which 
contains the prime number 73. 

That there was in the Maya system a number or order of units 
corresponding with Goodman's great cycle is certainly true, but this 
pertained to their numeral, and not their time, system. It is also 
admitted that the large quadruple glyph that usually heads the initial 
series is the symbol used to represent this number or order of units. 
But, as has been shown, there is no reason whatever for believing that 
they were numbered otherwise than in accordance with the vigesimal 
system; that is to say, 20 cycles to the great cycle, and 20 great cycles 
to the next higher unit. It is necessary, therefore, for Goodman, before 
his v theory can be accepted, to show by satisfactory evidence that, on 
reaching the cycles and great cycles, the ordinary method of proceed- 
ing by the vigesimal system was abandoned and other multiples were 
introduced. That there was a change from this rule in passing from 
the 2nd order of units, or chuens, to the 3rd order, or aha us, where 18 
was made the multiple, is proved by incontrovertible evidence and 
hence must be admitted, even though we ma} 7 not be able to show by 
absolute demonstration why the change was made. Nevertheless, we 
are justified in believing that, in this instance, the method of numera- 
tion was made to correspond with the number of months in the year. 
But no such reason appears for Goodman's proposed change in the 
higher orders of units; we are, therefore, justified in rejecting the 
idea until other proof, besides its necessity to support a theory, is 
shown. It must be made evident by proof that the series can not be 
otherwise explained, which we have shown is not the case, or it must 
be shown that the great cycle symbols present, by their forms, the 
numbers assigned them. 


Before referring to the numbers of the great cycles as obtained by a 
study of the forms of the symbols, I introduce the following quotation 
from Goodman's work (page 38) : 

The number and diversity of these signs and the fantastic character of some of 
them — notably the face series — suggest a hieratic design to conceal the purport 
of the inscriptions from the uninitiated; but I think the determinative feature of 
their numeration, the desire to give symmetry and grace to their glyphs, and the 
possible purpose to avoid sameness and repetition, sufficiently account for the 
variety without ascribing it to a cryptogramic intention. It is probable, there- 
fore, that all the other series of numerals were as intelligible to the populace as 
the simple one of dots and bars — being, as it were, a mere difference in the style 
of characters, such as is to be seen in fancy printing or ornamental sign- writing. 

While it is likely that in most instances there is a full series of similar signs, 
just enough modified to distinguish them from each other, running from 1 to 20, 
I do not think this to be the case throughout. It will be found, I believe, that 
there are many sporadic signs, or signs without any serial connection. The fre- 
quent use of certain numbers accounts for this, and it is to designate these that 
solitary symbols are oftenest employed. There will probably be more signs dis- 
covered for 13, 18, and 20, than for any other number. 

I do not claim that the value of any sign about to be given is correct beyond 
question. On the contrary, I think it very likely that in some instances I shall 
myself find reason for a change. But, as in most cases I shall explain why I have 
attached the value given to particular signs, the reader will not be misled, but 
can accept, reject, or modify my estimate, according to his own judgment. It 
will be only by persistent trial, assumption, alteration, and readjustment, until a 
figure that fulfils the requirement of every condition under which a character 
appears is hit upon, that we shall be able to fix the values of all the numeral 

That the great cycle symbol can be determined by position in a 
series, even though imperfect in form, is evident from what has been 
shown, but the number must be determined otherwise. In order to 
show on what Goodman bases his conclusion as to the numbers of the 
great cycles so far as determined by the form, I quote the following 
from his work (page 83) : 


Here the reckoning reverts to the 5-day period. It is multiplied by 72, making 
an ahau; that by 20, making a katun; that by 20 again, making a cycle; and that 
by 13, making a great cycle. The last multiplier is the outflaring trinal character 
at the top [figure 1 60] . It is a 13 sign, duplicated to balance the glyph. The two 
20 multipliers appear only in the first of the symbols given above — or, rather, 
only in that does the single one extend all the way to the bottom, as is commonly 
the case. There should be two separate signs, however, as shown in some of the 
glyphs; but I have selected these particular specimens for another purpose, 
which I shall presently state. The 20 sign in the first glyph looks like anything 
but the same sign in the other two, and resembles a fish more than anything else. 
Yet they are identical in character, both representing the feathered dragon, the 
fringed jaw alone of which, reduced to the cursive comb-like character, is 
the commonest sign for 20. The evolution of this character is so curious and 



[ETH. ANN. 22 

interesting that I herewith give a series of glyphs, all taken from great-cycle 
symbols, showing the gradations [figure 161]. 

The reason why I selected the particular symbols given above, is that I think the 
number of the great cycle is specifically stated in them. Close observers will 
have noticed several peculiar things about the great-cycle character. The most 
peculiar of these is that, while the form of the katun symbol is preserved in it 
fully in every other respect, the cauac sign disappears from the superfix and is 
replaced by some other character. In more than three-fourths of the dates in the 
54th great cycle a dragon's head occupies its place; a tiger's head predominates 
in the 55th, while the remainder is made up of faces and signs that may represent 
a day, a cycle, or some other period. Whatever their character, they have no 

Fig. 160. Great cycle symbols. Goodman, page.83. 

peculiarities that can at present be construed into numerals, except in case of the 
three glyphs here reproduced; so, if the others have any numeric value, it must 
be arbitrarily expressed. The three in question indicate the 54th great cycle, and 
I think that all of them announce that fact, but each in a different way. The 
center of the katun superfix in the first is composed of a sign for 18 and a face. 
If it were plainly the face for 3 we should be left in no doubt; but, in consequence 
of the defacement of the stone, it is impossible to determine if a band — the char- 
acteristic of the 3 head — extends across the forehead or not. In the second glyph 
the ik symbol — a sign for 6— appears in an inclosure that probably represents 9, 
but as the coil is not clearly discernible we are again left in uncertainty. The 
third glyph has the meaningless face, which elsewhere serves as a mere vehicle 

Fig. 161. Comb-like symbols for 20. Goodman, page 83. 

for numerals, bearing a sign for 9, surmounted by three objects evidently intended 
for spheres, whose value is doubled by the dotted lines in them, rendering it prob- 
able that the combination was designed to express 9 X 6 = 54. I make no claim 
to absolute certainty in any of these cases; but, however uncertain the renderings 
may be separately, they collectively derive a high degree of probability from a 
single significant fact. The unmistakable numeral sign in each glyph is a divisor 
of 54. That these glyphs — the only ones with recognizable numerals — should 
contain signs for three out of the six numbers by which 54 is divisible, is a circum- 
stance too singular to be attributed to accident when a more reasonable explana- 
tion is to be found in the theory that these three particular figures were chosen 
with the definite purpose of arriving at that number. 


As Goodman admits in the passage quoted, it is only in the three 
great cycle signs presented (see figure 100) that the evidence of num- 
bering is found ; let us examine this evidence. "Here," he says, "the 
reckoning returns to the 5-day period. It is multiplied by 72, making 
an ahau," yet he fails to allude to anything in the figure to justify the 
statement. That the comb-like characters and their substitutes have 
the value of 20 is probably correct, the sign being duplicated, as 
Goodman suggests, for the sake of symmetry. The fair inference is 
that in the katun symbol they indicate that this time period or order 
of units is equal to 20 ahaus (20x360=7,200). This admission, how- 
ever, as will be seen, is fatal to Goodman's theory. 

The three figures given represent, according to this author, the 
54th great cycle, and indicate by the details, but each in a different 
way, the number 54. This, he says, is shown in the first (a) in the 
center of the superfix, where he finds a sign of 18 and a face denoting 
3 — though he admits that the latter is too imperfect for positive deter- 
mination. The fact is that he has presented no proof that the dotted 
coil denotes 18. He asserts in his explanation of the ahau series on 
Stela J, Copan, copied in full in my previous paper, that the double coil 
denotes 18, but gives no proof to sustain the statement. His symbol 
for 18 in the ear ornament (page 87) is wholly different. Moreover, 
the face in the superfix, so far as the details remain, corresponds in 
no respect with the face numerals for 3 given on page 13 of his work, 
but on the contraiy bears a strong resemblance to at least two of the 
face characters for 1 (page 42). It is unnecessary to follow him in 
order to find the desired number in the other two figures (b, c), as not 
a particle- of proof is offered to sustain his assertions. It is apparent 
from his language that he felt his attempt here was a failure, but it 
was necessary to offer something on the point in behalf of his theory. 
Why 54 was given as the number of this great cycle, which begins 
with the day 4 Ahau 8 Cumhu, is apparent from the great-cycle 
column of his "Perpetual Chronological Calendar"; but his reason for 
beginning the series with 4 Ahau 13 Yax will be referred to farther on. 

In order to make clear what is stated below in regard to the forms 
and details of the symbols of the great cycle, katun, etc., a number of 
the types of the great-cycle symbol are shown in figure 1 62 ; of the ahau 
in figure 1G3; of the katun in figure 104; and of the cycle in figure 105. 

That this symbol — several varieties of which are shown in figure 
102 (also seen in figure 100, and as initials in plates lxxi-lxxiii, 
lxxv, lxxvi, and lxxviii, and figures 140, 147, 151, and 15-S)— is 
built up from, or based on, the 300-day or ahau symbol of 1 lie ordi- 
nary form, as shown in number 9, figure 103, is evident. The katun 
symbol of the ordinary type (k, figure 104), has the same body form 
as the ahau symbol, but there is added a superfix consisting of a 
comb-like figure on each side, with a middle character usually resem- 
bling a Cauae symbol. 



[ETH. ANN. 22 



Pig. 162. Types of great cycle symbols from the inscriptions. 1, Stela E, Copan, Maudslay, part 
3, plate xlix; 2, Stela I, Copan, Mandslay, part 3, plate lxv; 3, Stela D, Quirigua, west side, 
Maudslay, part 12, plate xxvi; 4, Stela D, Quirigua, east side, Maudslay, part 12, plate xxv; 
5, Stela J, Quirigua, Maudslay, part 12, plate xlvi; 0, Stela K, Quirigua, Maudslay, part 12, plate 
xlix; 7, Monolithic Animal B, Quirigua, Maudslay, part 11, plate xiv; X and 9, Stela C, Copan, 
Maudslay, part 2, plate xli (both specimens on this plate t; 111, Stela A, Copan, Maudslay, part 2, 
plate xxx; 11, Altar S, Copan, Maudslay, pai-t 4, plate xciv; L2, Stela N, Copan, east side, Mauds- 
lay, part 4, plate LXXIX; 115. Stela N", Copan, west side, Maudslay, part 4, plate LXXIX, glyph 14, 
counting from the top; 14, supposed great-cycle symbol irom the Dresden codex, plate XLlll. 




This is an evident approach to the great-cycle symbol, as may be seen 
by comparing number 9, figure 163, with the types of the usual form 
shown in numbers 1 and 2, figure 162. The usual cycle symbol or 
symbol of the 5th order of units (figure 165 and figure 118) does not 
follow the ahau type, being wholly different in form. But an exam- 
ination of the great-cycle symbols given in numbers 1 and 2, figure 162, 
and in the other figures referred to above shows clearly that they are 
based on the ahau symbol. If the additions to the ahau symbol in 
order to form this symbol have any number signification — and it is 
reasonable to suppose that they do, as the symbols are numeral 
characters — then Goodman is probably right in assuming that the 
comb-like figures (the center character being variable) denote 20 as a 
multiple. The ordinary cycle symbol varies from the ahau type, 
being made up of two Cauac characters; but these have the same sig- 

Fig. 163. Types of the ahau (360) symbol. 

nification, if Goodman be right, as the comb-like figures in the katun 
and great-cycle symbols — that is, 20. Of this, however, we have no 
positive proof, except it be found in the symbol itself, where the char- 
acter is, or the two combined are, beyond question, used to represent 
a number. An examination of the face characters for this period or 
order of units shows that, as a general rule, the symbol of 20 or full 
count (equals 0) (see figure 111) is present in the form of a hand 
across the lower jaw. We have also called attention to the fact that 
the only face character of the great cycle found in the inscriptions 
(see glyph 5, figure 158) lias the hand across the lower jaw, indicating 
that it is equivalent to 20 of the next lower order, thai is, 20 cycles. 

There is, in fact, seemingly positive eyide nee thai the superfix of 
the great-cycle characters does not and can not give the number 54, as 
those which represent this great cycle, be its number whal it may, 
differ from one another, as will he seen by reference to figure L62, 



[ETH. ANN. 22 

numbers 2 to 12. Having worked out his system in tabular form, 
Goodman finds that 4 Ahau 8 Cumhu is the first day of his 54th great 

( Fig. 164. Types of the katun symbol. 

cycle, assuming, as he does, that 4 Ahau 13 Yax was the first day of 
his grand era. The particular process by which he reached the con 



e / 9 

Fig. 165. Types of the cycle symbol. 

elusion that 4 Ahau 13 Yax was the initial day of his first great cycle, 
and hence of his grand era, is not clear. The choice was apparently 


arbitrary, though it was necessary that the date chosen should make 
connection with -I Ahau 8 Cumliu as the first day of a great cycle. 
His explanation of the grand era, on pages 26 and 27 of his work, 
shows the relation of the minor periods to it according to his theory, 
but does not give the reason for selecting 4 Ahau 13 Yax as the initial 
date. On page 34 he speaks of the date as an important one in the 
inscriptions, but still does not give the reason for making it the 
beginning of the grand era. 

That any other 4 Ahau, which would bring 4 Ahau 8 Cumhu as the 
first day of a great cycle, would answer as well as 4 Ahau 13 Yax, 
even on his theory, is easily shown. As the Mayan time count is an 
orderly round, a given day recurring at the end of a certain period, it 
is evident, as eveiyone acquainted with the system knows, that the 
count of periods may begin at any point, unless some fixed point in 
the series is found with its proper number. One check in this respect 
found in the inscriptions is the fact just mentioned that, according 
to Goodman's system, 4 Ahau 8 Cumhu appears to be the initial day 
of a great cycle, and the initial dates of the other great cycles must 
fit correctly with this determined initial date — that is to say, following 
his theory and counting 13 cycles to the great cycle, these initial dates 
must all be a day 4 Ahau. Another possible check is the long series 
in the Copan inscription, which goes back 14 great cycles preceding 
that beginning with 4 Ahau 8 Cumhu. 

Let us turn to Goodman's "Perpetual Chronological Calendar," to 
the great-cycle column. Suppose that instead of commencing with 
the date 4 Ahau 13 Yax, with which he begins the grand era, we begin 
with 4 Ahau 18 Zotz, the initial day of his 40th great cycle. The series 
will then be as follows, if we adopt his method of numbering: 


4 Ahau 18 Zotz 


4 Ahau 18 Cumhu 


4 Ahau 13 Kankin 


4 Ahau 8 Yax 


4 Ahau 3 Xul 


4 Ahau 18 Pop 


4 Ahau 18 Muan 


4 Ahau 13 Zac 


4 Ahau 8 Yaxkin 


4 Ahau 3 Zip 


4 Ahau 3 Kayab 


4 Ahau 18 Ceh 


4 Ahau 13 Mol 


4 Ahau 8 Zotz 


4 Ahau 8 Cumhu 


4 Ahau 3 Kankin 


4 Ahau 18 Chen 


4 Ahau 13 Tzec 


4 Ahau 8 Pop 


4 Ahau 8 Muan 


4 Ahau 3 Zac 


4 Ahau 18 Xul 


4 Ahau 13 Uo 


4 Ahau 13 Pax 


4 Ahau 8 Ceh 


4 Ahau 3 Mol 


4 Ahau 18 Zip 


4 Ahau 18 Kayab 


4 Ahau 13 Mac 


4 Ahau 8 Chen 


4 Ahau 3 Tzec 


4 Ahau 3 Uayeb 


4 Ahau 18 Kankin 


4 Ahau 13 Yax 


4 Ahau 8 Xul 

And so on to tin 1 72d, the next being 4 Ahau 18 Zotz, with which 
the numbering began. 

270 MAYAN CALENDAR SYSTEMS [eth. ann. 22 

This will meet every requirement, including the limitations above 
mentioned, as fully and as completely as the series given by Goodman, 
even if we hold to his theory of 13 cycles to the great cycle and 
73 great cycles to his grand era, and follow his own method of 
counting. The same thing is true if we select, as the first great cycle, 
any other of the 40 which precede that with which we began the count. 

There is another fact which appears to conflict with Goodman's 
theory and, indeed, to be irreconcilable with it. According to this 
theory, the grand era, consisting of 136,656,000 days, is the least 
common multiple of all the different factors of the regular calendar 
as well as of his chronological calendar, at the beginning of which 
all the periods start anew on their grand round. That this number 
is the common multiple of all these periods or factors is true. But 
how are we to reconcile the theory with the fact that he begins this 
great era with the day 4 Ahau 13 Yax, which is certainly not the 
beginning day of a year or of a month? It is true the 136,656,000 
days is an exact multiple of 365, but, starting the count of 365 with 
the day 4 Ahau 13 Yax makes the latter number a mere numeral 
factor; no regular Mayan year could begin with the day 4 Ahau or 
with the 13th day of the month Yax. From February 1, 1899, to the 
following January 31, in our time system, is a year's time, but the 
period is composed of parts of two calendar years. 

Goodman's theory, in order to be correct and keep the time periods 
in proper order, if his grand era is a true and absolute rounding-out 
period of all the minor periods, absolutely requires that this great 
period shall begin with the 1st day (or 20th if he prefers this number- 
ing) of the month Pop, and the first year of the 52-year cycle or calen- 
dar round. Otherwise, when the era ends, it will be in the middle 
of a year, as it will if it begins on 4 Ahau 13 Yax, and closes with 
3 Cauac 12 Yax. 

The question next in importance is, are his tables correct, though 
based on an erroneous theory? Those of the first series, termed the 
"Archaic Annual Calendar," are nothing more than the ordinary cal- 
endars of the 52 years of what has heretofore been termed a "cycle," 
but to which he applies the name " calendar round," each year being 
given separately. They are all contained in my condensed calendar. 
This is nothing new, as the method had been in use for a number of 
years before Goodman commenced his investigations. As his ' 'Archaic 
Chronological Calendar" is nothing more than a continuous series of 
ahaus, or 360-day periods, using Ahau as the " initial day " through 39 
of the 5th order of units, following one another in regular succession, 
it is correct — with certain exceptions to be noted — where Ahau is used 
as the initial da} 7 in the count, but will not apply when any other day 
is selected as the initial date. It is erroneous in counting 13 of the 
cycles or the 5th order of units to the next higher order, and in begin- 
ning the numbering of the so-called periods with 73, 13, and 20. His 
tables of years are also erroneous in the latter respect. 


It is apparent to anyone at all acquainted with the Mayan time and 
numeral systems that, having a continuous series of days written out 
in regular order and of sufficient length, with the day numbers and 
month numbers attached, we may start at any point and count off the 
numbers given in the ahau, katun, and cycle periods, and we will 
have precisely what is given in Goodman's "Archaic Chronological 
Calendar," except that we may have some other initial day than 
Ahau. If it should be Kan it would at some point correspond exactly 
with the series of the Dresden codex which have been referred to; if 
Ahau, then the periods would agree with those of the inscriptions and 
some of those in the Dresden codex. Now, it is evident that in count- 
ing off a number in the next higher group above the so-called cycle, 
if we count off the latter periods by 20, instead of 13, the succession 
would be as regular as in the other case, there being nothing whatever 
in the system requiring or even suggesting 13. Hence we might take 
Goodman's tables, if more extended, and making 4 Ahau 8 Cumhu 
our starting point, count forward or backward by steps of 20 cycles 
each, and thus find the correct initial days of the great cycles as we 
have shown above. With the tables given in his work we can only 
count forward from the beginning of his 54th great cycle to the 7th 
cycle of the 55th great cycle as he has numbered them, showing that 
10 Ahau 13 Yaxkin is the beginning day of the next great cycle, 
counting 20 cycles to the great cycle, which I have shown to be the 
correct method. 

I shall not discuss Goodman's theory of the number values of the 
day and month symbols, as there does not appear to have been any 
use made of them as numerals. 

Let us turn again to the order in which the numbers of the ahaus 
follow one another, to wit: 13, 9, 5, 1, 10, 6, 2, 11, 7, 3, 12, 8, 4, 13, 
etc. This has been fully discussed in one light in this paper, but the 
object at present is to view it in another light and with special refer- 
ence to Goodman's theory in regard to it. That has also been 
noticed to some extent in my previous paper, but there are some 
points omitted in that discussion to which it is desirable to call atten- 
tion. I quote in full Goodman's statement of his discovery of the 
order of succession : 

Ymix is the day following Ahau; hence, I reasoned to myself, if a period begin 
with the former it must terminate with the latter: moreover, 1 succeeding 13 in 
the day count, if 1 Ymix begin a period 13 Ahau must end it; and, further, this 
period being composed of 18 lesser ones of 20 years each, it is at a distance of 260 
years apart in the annual calendar that I must look for a corresponding 1 Ymix 
and 13 Ahau, recollecting that I need not expect to find them falling on any fixed 
date. But, as the order of the 13 subdi visions is given, with the terminal Ahau 
numbers, it is not necessary to attempt so extended a research, and prudence dic- 
tates that I keep my experiments within the narrowest possible limits to guard 
against mistake. I will, therefore, at the start proceed only to the end of the first 
20-year period, or katun, and look for 11 Ahau. The trial is made. It proves 
abortive, as I anticii>ated. The Ahau number at the end of 20 years is 7 instead 
of 11. The desired 11 Ahau is 5 months away to the left. It is the same old 

272 MAYAN CALENDAE SYSTEMS [eth. ann. 22 

story of failure over again. But wait a minute! Five months are equivalent to 
100 days. To divide by 20 would take just 5 days from each of the 20 years of 
the katun. Years? What if they were not years at all that Landa was talking 
about, but only periods of 360 days? They may be the ahaus. Let me hasten to 
find out how the numbers will run in a division of this possible Katun into 20 
such periods. Here it is: 9, 5, 1," 10, 6, 2, 11, 7, 3, 12, 8, 4, 13, 9, 5, 1, 10, 6, 2, 11. 
Ah, this is significant! That paragraph of Perez, what are its exact words? 
"The Indians of Yucatan had yet another species of cycle, but as the method 
followed by them in using it can not be found, nor any example by which an 
idea of its nature might be imagined, I shall only copy what is literally said of it 
in a manuscript, viz: ' There was another number which they called ua katun, 
and which served them as a key to find the katuns. According to the order of 
its march it falls on the days of the uayeb yaab and revolves to the end of cer- 
tain years: katunes 13, 9, 5, 1, 10, 6, 2, 11, 7, 3, 12, 8, 4.' " Poor Don Pio! To have 
the pearl in his grasp and be unaware of its pricelessness, like so many others. 
But I must not exult too much yet. The succession of the katuns, reckoned 
according to this principle, is yet to be ascertained before my fancied discovery 
can be established by a crucial test. I score the ahaus off in the foregoing order, 
and, sure enough, the 20ths give the desired result, 11, 9, 7, 5, 3, 1, 12, 10, 8, 6, 4, 
2, 13. Eureka! The perturbed spirit of the Maya calendar, which has endeav- 
ored so long to impart its message to the world, may rest at last. 

That taking the day numbers of the first days of the ahaus in a 
katun will give the order of succession mentioned is certainly true, as 
we have shown, but the question to be discussed here relates to the 
statement of the authority quoted by Perez. According to this state- 
ment as given by Goodman, " There was another number which they 
called ua katun, and which served them as a key to find the katuns. 
According to the order of its march it falls on the days of the uayeb 
vaab, and revolves to the end of certain years; katunes 13, 9, 5, 1, 10, 
6, 2, 11, 7, 3, 12, 8, 4." 

It will be best, however, to give Perez's exact words as found in the 
appendix to Brasseur's edition of Landa's " De las Cosas," page 418: 

' ' Habia otro numero que llamaban Ua Katun el que les servia como 11a va para 
acertar y hallar los katunes, y segun el orden de sus movimientos cae a los dos 
dias del Uayeb haab y da su vuelta al cabo de algunos anos: Katunes 13, 9, 5, 1, 10, 
6,2,11,-7. 3, 12,8,4." 

Brasseur's translation is as follows : 

' ' lis avaient un autre chiffre qirils appelaient Ua Katun, qui leur servait comme 
de clef, pour ajuster et trouver les katun et suivant l'ordre de ses mouvements, il 
tombe aux deux jours du Uayab haab et retourne a la fin de quelques annees: Katun 
13, 9, 5, 1, 10, 6, 2, 11, 7, 3, 12, 8, 4." 

A closer translation than that by Goodman, which omits one impor- 
tant word, may be given as follows : 

They have another number which they called ua katun, which served them 
as a key to regulate and find the katuns, and according to the order of its move- 
ments falls on the two days of the uayeb haab and returns at the end of certain 
years; katuns 13, etc. 

The important word omitted by Goodman and which is usually 
omitted in English translations is the "two." Brasseur's translation 

thomas] THE SUCCESSION 13, 9, 5, 1, 10, ETC. 273 

contains it, and Perez recognizes it by his (erroneous) reference on 
the same page as the passage quoted, the "second." intercalary day. 
I called special attention to this important word in my "Stud} 7 of the 
Manuscript Troano," page 55. 

Now, it is certain that the unknown author of this passage was 
somewhat familiar with the Maya time system, otherwise he could, 
not have hit upon this order of numbers which is found in at least 
three different relations in the system; and it is also certain that his 
reference is to true Mayan years (as is shown by the reference to the 
uayeb haab, or five intercalated days), and can not be made to apply 
to Goodman's ahaus. 

As the term "years" in the passage quoted, can have no other pos- 
sible meaning than that of 365 days, the question arises, what is 
meant by the term "katun" as therein used? That it could not be 
Goodman's katun of 7,200 days, or 20 ahaus of 360 days each (which 
Seler also claims to have discovered) , is evident. Although we may not 
be able to demonstrate what is meant by the term in this connection, 
we can show where and how this order of succession occurs, using the 
last of the intercalated days. As the number of the day with which 
the }^ear ends is the same as that with which it begins, the order will 
be precisely the same as that in which the years are numbered. If 
the calendar of the inscriptions and the Dresden codex is used, whose 
dominical days are Akbal, Lamat, Ben, and Ezanab, the terminal days 
will be Manik, Eb, Caban, and Ik, and their numbers in the successive 
years will be as shown in the following table, which extends through 
the cycle of 52 years, after and before which the same series will be 
repeated : • 

Manik Eb Caban Ik 




















































Beginning at the bottom and running up the right-hand column, we 
find precisely the order of succession given in the quotation, to wit, 
13, 9, 5, 1, 10, 6, etc. Precisely the same order will be found by run- 
ning up either of the other columns. Kach step, it is true, covers four 
years, but it forms a basis for easy and ready counting; moreover, 
the quotation says, "returns at the end of certain years." It does 

22 eth— 04 18 

274 MAYAN CALENDAR SYSTEMS [eth. ann. 22 

not say the "numbers" which so return are katuns, but that they 
"served as a key to find the katuns," clearly distinguishing between 
the "katuns" and "certain years." There is nothing, therefore, in 
the quotation which implies that the numbers in the series 13, 9, 5, etc., 
were the numbers of the katuns, nor is there any mention therein of 
the numbers of the katuns or of the number of years constituting a 
katun. It is to Landa that we must go for information on the latter 
point. According to his statement, which has been oft repeated, the 
Mayas counted their ages by 20 3^ears, a but he says nothing in reference 
to the order of numbering. h 

As the periods referred to are unquestionably years, the katuns 
must be periods of years; and writers who have so contended are cor- 
rect in this respect, and 20 years is the number assigned to a katun by 
all the early authorities, whether right or wrong. 

The direction of counting, it is true, is backward, but, as Good- 
man states, the reference among the Mayas is generally to past time, 
and the example Landa gives of counting time, in connection with 
the passage referred to, relates to what had passed. He says an 
elderly man of whom he had spoken could easily count back 300 years 
by means of the katuns or ages. This author, if I rightly understand 
his language, indicates that they had a still higher count of 13 x 20 
years. His language is as follows : 

No solo tenian los indios cuenta en el aiio y meses, como queda dicho, y seiia- 
lado atras pero tenian cierto modo de contar los tiempos y sus cosas por edades, 
las quales hazian de veynte en veynte anos, contando xiii veyntes con una de las 
xx letras de los meses que Hainan Ahau, sin orden sino retruecanados como 
pareceran en la siguiente raya redonda; Hainan les a estos en su lengua Katunes. a 

Thirteen times 20 is 260, or five cycles of 52 years each, the same num- 
ber of 3 T ears that there are days in their so-called sacred year. Possi- 
bly, however, he may refer here to the 260-day period. 

When we free our minds entirely from any thought that ahaus, 
katuns, etc. , represent or have any relation to time periods, and look 
upon them merely as numbers, just as we think of tens, hundreds, etc., 
the difficulties raised by Goodman's theory of a Maya " chronological 
calendar " vanish. The Mayas of one section, for some historical, tradi- 
tional, or mythological reason, selected a particular initial date for 
their era, and, as a usual thing, counted long periods from it, and in 
doing so used numbers in accordance with their numeral system, and 
represented these in their inscriptions by certain symbols. This is 
all of Goodman's supposed wonderful chronological system — this and 
nothing more. 

It would have been much better if he had used the real Mayan 
numeral terms as they stand (as Dr Brinton has suggested), or in a 

a Landa, De Las Cosas, p. 312. 

bit will doubtless be recalled that in the " Study of the Manuscript T^ano" I contended that 
the ahaus or katuns consisted of 24 years, basing my conclusion on the order given above; but 
a more careful study of the passage quoted above from Perez does not necessarily indicate that 
these periods were numbered according to the order given. 


modified form, to indicate the variation of time numeration from the 
regular vigesimal system, thus: 

20 units = 1 kal in place of chuen. 

18 kal = 1 bak in place of ahau. 

20 bak = 1 pic in place of katun. 

20 pic = 1 calab in place of cycle. 

20 calab = 1 kinchil in place of great cycle. 

20 kinchil = 1 ahau in place of grand era. 

It is true that above the kal the numbers would vary from the 
true vigesimal count in consequence of counting but 18 instead of 20 
kal to the next higher order. This, however, might have been shown 
by prefixing "minor," thus, "minor bak," "minor pic," etc., but no 
real confusion would have resulted from using the simple names as 
Brinton has suggested. Seler suggests "uinal" in place of chuen; 
"tun " in place of ahau, but retains "katun " as applied by Goodman. 


If the "Annals of the Cakchiquels," written or supposed to have 
been written soon after the Spanish conquest by a member of the 
Xahila family, are to be trusted in regard to the Cakchiquel calendar 
system, this system was peculiar, differing in some important respects 
from that of the JVIayas, which has been described in the preceding 
part of this paper. All that is known in regard to its peculiar features 
is found in these Annals, and must be gathered from incidental men- 
tion of dates. In order to place the data before the reader, I quote the 
more important of these mentions from the translation by Dv Brinton 
in the Library of Aboriginal American Literature, vi, "The Annals 
of the Cakchiquels," 1885. 

As a noted revolt, described as the "revolt at Iximche," is selected 
by the author of the Annals as the era from which to reckon all sub- 
sequent events, we begin the quotations with the passages referring 
to and fixing the date of this event. 

(1) The day of the revolt was appointed by this chief, Cay Hunahpu, and on 
this day, 11th Ah, the revolt broke out [page 157]. 

(2) Thirty-one days after the revolt, as the Quiches desired to destroy those 
of Tibaqoy, these Tukuches removed to Chiavar and put to death the Quiches, 
who yielded in a battle at a place named Yaxontzui, on the day 9th Caok [page 

(3) On the 36th day after the revolt Cinahitoh perished ... on the day 11th 
Can [ibid.]. 

(4) One year less ten days after the revolt was hanged the chief orator Ahmox- 
nay on the day 11th Akbal [ibid.]. 

(5) The day 8 Ah was one year after the Revolt [page 161]. 

(6) The day 5 Ah was two years after the Revolt [ibid.]. 
(T) The day 2 Ah was three years after the Revolt [ibid. ] . 

(8) The day 12 Ah completed the fourth year after the Revolt [ibid.]. 

(9) The 9 Ah completed the fifth year after the Revolt [page 163]. 

(10) The 6 Ah completed the sixth year after the Revolt [ibid.]. 

(11) On the 3 Ah there were seven years from the Revolt [ibid.]. 

27() MAYAN CALENDAR SYSTEMS [eth. ann. 22 

(12) In the eighth year after the revolt, the Tzutuhils were defeated by those 
of Xeynup and Xepalica; they were slaughtered, Zakbin and Ahmak having 
perished in the action on the day 13 Ahmak [ibid.]. 

(13) On the day 13 Ah there were eight years from the revolt [ibid.]. 

(14) On 10 Ah there were nine years from the revolt [ibid.]. 

(15) Twelve days were lacking to complete the tenth year after the revolt 
. . . the day 8 Imox [ibid.]. 

(16) The day 7 Ah completed the tenth year after the Revolt [ibid.] . 

(17) On 4 Ah there were eleven years after the Revolt [ibid.]. 

(18) On 1 Ah there were twelve years [ibid.]. 

(19) On 11 Ah there were thirteen years after the Revolt [ibid.]; 

(20) The day 8 Ah completed the 14th year after the Revolt [page 165] . 

(21) The day 5 Ah completed the 15th year after the Revolt [ibid.]. 

(22) The day 2 Ah completed the 16th year after the Revolt [ibid.]. 

(23) The day 12 Ah completed the 17th year after the Revolt [page 167]. 

(24) The day 9 Ah completed the 18th year after the Revolt [ibid.] . • 

(25) On the day 3 Caok the doves passed over the city of Iximche. . . . One 
hundred days after the doves had been seen the locusts came ... on the day 2 
Yg [ibid.]. ' 

(26) The day 8 [6?] Ah completed the 19th year after the Revolt [ibid.] 

(27) The day 3 Ah completed one cycle [page 169]. 

(28) With the day 13 Ah another year was completed [ibid.] . 

(29) A second year was completed on the day 10 Ah, after the Revolt [ibid.]. 

(30) On the day 7 Ah was completed the third year of the second cycle after 
the Revolt [ibid.]. 

So far the dates given are in regular succession as found in the 
Annals; the others given are only those which are considered impor- 

(31) On the day 14 [12?] Carney died the King Hunyg [page 171]. ... A 
hundred days after the death of the kings Hunyg and Lahuh Noh, there were 
elected as kings Cahi Imox and Belehe Qat, on the day 1 Can [page 173]. 

(3,2) Twenty days after the chiefs began to rule there was an insurrection 
... on the day 10 Queh [page 175]. 

(33) We married your mother, O my children, one year after the death of 
your grandfather [Hunyg]. We took her to wife on the day 12 Toh [pages 

(34) On the day 5 Ah was the eighth year of the first [second] cycle. It was 
during this year [meaning the year following?] that the Castilians arrived. . . . 

On the day 1 Ganal the Quiches were destroyed by the Castilians On the 

day 4 Qat three chiefs, the king and the next in rank, were burned alive by 
Tunatiuh [page 177] . 

(35) It was on the day 1 Hunahpu when the Castilians arrived at Iximche 
with their chief, Tunatiuh. . . . Only five days after, Tunatiuh went forth from 
the capital. Then the Tzutuhils were conquered by the Castilians. It was the 
day 7 Carney [page 179]. 

(36) Twenty- five days afterwards Tunatiuh went forth from the capital to 
Cuzcatan . . . On the day 2 Queh Atacat was slain . . . On the day 10 
Hunahpu he [Tunatiuh] returned from Cuzcatan. He had been absent only 40 
days [page 181]. 

(37) Our city [Iximche] was abandoned on the day 7 Amak . . . Ten days after 
we had left the city, war was begun by Tunatiuh ... on the day 4 Carney . . . 
One hundred and eighty days after the desertion of the city was completed the 
ninth year (of the second cycle). On the day 2 Ah was completed the 29th year 
after the Revolt [page 183] . 


(38) There were lacking 120 days to complete two years since we had abandoned 
the capital when Tunatinh came there in order to set fire to the city. On the day 
4 Carney, two years less six months after the beginning of the war, he set fire to 
the capital and returned [page 185] . 

(39) On the day 12 Ah was completed the 30th year after the Revolt [ibid.]. 

(40) On the day 9 Ah was completed the 31st year after the Revolt [ibid.]. 

(41) In the course of the following year . . . Chiixot was abandoned. . . . 
Three hundred days after Chiixot was taken began the payment of tribute 
... on the day 6 Tzi [pages 185-187]. 

(42) It was two years less 120 days after the beginning of the tribute when 
died the chief Ahtun cue Tihax ... on the day 6 Akbal. . . . On the day 3 Ah 
was completed the 33d year [page 187] . 

(43) For 86 days these chiefs had hid in the woods. . . . On the day 7 Ahmak 
the chiefs decided to come forth. . . . On the day 13 Ah was completed the 36th 
[34th] year after the Revolt [page 187] . 

(44) On the 10th Ah was completed the 35th year after the Revolt. Forty days 
were lacking to complete three years from the date of the submission of the kings 
when Belehe Qat died ... on the 7th Queh [page 188] . 

(45) On the 8th Ah was completed the 40th year after the revolt. On the 5th 
Ah was completed the first year of the third cycle [page 189] . 

(46) It was on the day 11 Ahmak that he [Tunatiuh] killed the Ah-tzib. On the 
day 2 Ah was completed the second year of the third cycle. One hundred and 
twenty days after the death of Ahtzib and of the return of Tunatiuh. the prince 
Mantunalo departed . . . Two hundred and sixty days after his return Tunatiuh 
hanged the king Ahpozotzil Cahi Imox, on the day 13 Ganel [pages 189-190]. 

(47) The day 12 Ah completed the third year of the third cycle. Two hun- 
dred and eighty days after the execution of the king Ahpozotzil he hanged Chuvy 
Tziquinu . . . on the day 4 Can [page 190] . 

(48) On the day 9 Ah was completed the fourth year of the third cycle after the 
revolt. . . . On the day 2 Tihax . . . the wife of Tunatiuh was drowned. One 
hundred and sixty days after this disaster there arrived our fathers of St. Dominic 
... on the day 12 Batz [page 190] . 

(49) On the day 8 Ah was completed the 13th year of the third cycle. . , . 
Ahtzil Juan Perez . . . died on the day 12 Tihax. Eighty days after . . . there 
was an eruption of the mountain Chigag ... on the day 9 Ah . . . On the day 
12 Ah was completed the 16th year of the third cycle [page 192]. 

(50) Died the chief Don Francisco Ahpozotzil . . . on the day 1 Can, a Monday, 
the 14th day of the month October [page 193] . 

(51) On the day 6 Ah was completed the 18th year of the third cycle. ... In 
the 13th month the day of Sanctiago occurred on the day 1 Tziquin. . . . On that 
day was inaugurated . . . the Emperor Don Peliphe. . . . The day St. Francis 
[was] the day 7 Carney [pages 193-194]. 

(52) On the day 3 Ah was completed the 19th year of the third cycle after the 
revolt. The Alcaldes in the year 1557 were . . . The day 5 Ey [was] 20 days 
before the close of the third cycle. ... On the day 13 Ah was completed the 
third cycle ... in the year 1558 [page 194]. 

The foregoing notes and quotations contain, it is believed, all the 
data found in the "Annals" throwing any light on the Cakchiquel 
calendar. But in order that the reader, who may not have the works 
relating to this calendar at hand, may be furnished with the data 
necessary to follow me in my discussion, I introduce here a List of 
the days of this calendar in the, order usually given, with those of the 
Maya calendar placed beside them in corresponding order. 



[ETH. ANN. 22 

Days of the Cakchiquel and Maya Calendars 

Cakchiquel Maya Cakchiquel Maya 

days days days days 

1 Imox Imix 11 Batz Chuen 

2 Ig or Yg Ik , 12 Ee Eb 

3 Akbal Akbal 13 Ah Ben 

4 Kat Kan 14 Yiz Ix 

5 Can Chicchan 15 Tziquin Men 

6 Carney Cimi 16 Ahmak Cib 

7 Qneh Manik 17 Noli Caban 

8 Kanel Lamat 18 Tihax Ezanab 

9 Toh Mulnc 19 Caok Canac 
10 Tzii Oc 20 Hunahpu Aliau 

As the author of the Annals ends the year with the day Ah, it 
must have begun with Yiz, if there was no arbitrary change in the 
succession of days. The following condensed calendar is therefore 
constructed on this basis as a means of counting time : 

Cakchiquel Calendar 

1 Yiz :__,._ 1 8 2 9 3 10 4 11 . 5 12 6 13 7 

2 Tziquin 2 9 3 10 4 11 5 12 6 13 7 1 8 

3 Ahmak 3 10 4 11 5 12 6 13 7 1 8 2 9 

4 Noh 4 11 5 12 6 13 7 1 8 2 9 3 10 

5 Tihax 5 12 6 13 7 1 8 2 9 3 10 4 11 

6 Caok 6 13 7 1 8 2 9 3 10 4 11 5 12 

7 Hunahpu 7 18 29 3 10 4 11 5 12 6 13 

8 Imox 8 2 9 3 10 4 11 5 12 6 13 7 1 

9 Ik 9 3 10 4 11 5 12 6 13 7 1 8 2 

10 Akbal 10 4 11 5 12 6 13 7 1 8 2 9 3 

11 Kat 11 5 12 6 13 7 1 8 2 9 3 10 4 

12 Can 12 6 13 7 1 8 2 9 3 10 4 11 5 

13Caniey 13 7 1 8 2 9 3 10 4 11 5 12 6 

14Queh 1 8 2 9 3 10 4 11 5 12 6 13 7 

15Kanel 2 9 3 10 4 11 5 12 6 13 7 1 8 

16 Toh 3 10 4 11 5 12 6 13 7 1 8 2 9 

17 Tzii 4 11 5 12 6 13 7 1 8 2 9 3 10 

18Batz__: 5 12 6 13 7 1 8 2 9 3 10 4 11 

19 Ee..'.. 6 13 7 1 8 2 9 3 10 4 11 5 12 

20Ah 7 1 8 2 9 3 10 4 11 5 12' 6 13 

In using this to count forward, we count on to the end of the right- 
hand column and then go back to the left-hand column. To count 
backward, the direction is reversed. 

It will be observed from the quotations given that the 3-ears all end 
with the day Ah, that the numbering of the days is by 1 to 13 as usual, 
and that the terminal Ahs of the years succeed one another in the 
following order: 11 Ah, 8 Ah, 5 Ah, etc., giving the descending series 
11, 8, 5, 2, 12, 9, 6, 3, 13, 10, 7, 4, 1, 11, 8, etc., the number of any 
given year being 3 less than that of the one which preceded. 

It is apparent, therefore, that the year could not have consisted of 


365 days, that is, of 18 months of 20 days each and 5 added days, for 

even the supposition that these added days were neither numbered 

nor counted does not give the order found in the Annals. Nor will 

Goodman's supposition that they counted 366 days to the year give 

this succession, though he counts the system alluded to in the 

Annals as distinct from the Cakchiquel Annual Calendar. Brinton 

says : 

The calendars in use were of two different kinds, the one called qliol kill, lit- 
erally ''the valuer or appraiser of days," which was employed exclusively for 
astrological and divining purposes, to decide on which were lucky and unlucky 
days, and may kill, " the revolution or recurrence of days," which was for chrono- 
logical purposes." 

I find no other explanation of a calendar which would end in the 
manner mentioned in the Annals, than a year of 20 months of 20 da} T s 
each, or 400 days, the days being numbered in the usual Mayan 
method of 1 to 13. Seler* gives this explanation and Goodman also 
adopts it for their chronological year. That if we count this num- 
ber of months to the year the different years will end on the same 
day is evident, and that the day numbers will follow one another in 
the order mentioned above can be seen by reference to the above 
condensed calendar. If we count 20 months, the year beginning with 
1 Yiz will end with 10 Ah, and the next year will begin with 11 Yiz; 
or if we commence with the column headed 11, and count 20 months, 
the year will end with 7 Ah, and the next year will begin with 8 Yiz; 
if we commence with the column headed 8, and count 20 months, the 
year will end with 4 Ah, etc. This appears to be the only explanation 
of this singular calendar, if we suppose the annalist to be correct in 
his statements as to the dates on which the years ended. 

As proof that the annalist counted 400 days to the year we have 
the following evidence from the above quotations: By number 1, we 
learn that the Revolt, which he takes as the beginning of his era, took 
place on 11 Ah; by number 5 we see that the first year of the Revolt 
ended on 8 Ah; in number 4 it is stated that "One year less ten days 
after the revolt was hanged the chief orator Ahmoxnoy, on the day 11 
Akbal." The day 11 Akbal will occur twice only in the ordinary } T ear 
of 365 days, and twice only in the year of 400 days. As the Revolt 
occurred on 11 Ah, the first year thereafter must have begun with 
the day 12 Yiz. The day 11 Akbal would occur first at the end of 6 
months and 10 days — or 130 days. That 10 days added to this could 
not have completed the year will be conceded. The next occurrence 
of 11 Akbal would be at the end of 19 months and 10 days, or 390 
days, 10 days more reaching the day 8 Ah, the end of the first year. 
Although neither 140 nor 400 days correspond will) any natural 
phenomena it is safe to assume that 400 days was the period the 
annalist referred to and not 140 days. 

"Annals of the Cakchiquels, Philadelphia, 1885, p. 31. 

b Transactions Berlin Anthropological Society, June, 1889. 

280 MAYAN CALENDAR SYSTEMS [eth. ann. 22 

In number 15 it is stated that 12 clays were lacking to complete the 
tenth year after the Revolt, etc., on the day 8 Imox; and in number 
10 that the day 7 Ah completed the 10th year. As it is stated in 
number 14 that 10 Ah was the end of the 9th year after the Revolt, 8 
Imox would occur 128 and t>88 days thereafter. Counting 12 days 
from the latter brings us to 7 Ah and gives 400 as the number of days 
in the year. This result must be accepted, or we must decide that 
the year consisted of only 140 days, which is unreasonable. In num- 
ber 24 it is stated that 9 Ah completed the 18th year after the Revolt, 
and in number 26 that 8 (?) Ah completed the 19th year (that this 
should be 6 Ah is evident, as 9 Ah precedes and 3 Ah (number 27) 
follows it). In number 25, which relates to the 19th year, it is stated 
that on the day 3 Caok the doves passed over the city of Iximche; 
and that 100 days after the doves had been seen the locusts came, on 
the day 2 Yg (or Ik). Now, the first occurrence of 3 Caok in the 19th 
year after the Revolt, that is, the year following 9 Ah (the year begin- 
ning with 10 Yiz), is 2 months and 6 days after the commencement 
of the year. One hundred days more bring us to 12 Caok, the 6th 
day of the 8th month, or 7 months and 6 days from the commence- 
ment of the year. This is not the day given, but counting 4 days 
more we reach 2 Yg or Ik, the day named. As 1 00 is a round num- 
ber, the 1 04 may be assumed as correct. As this, even if the number 
be limited to 100, gives more than 140 days in this year we have evi- 
dence that a year of 400 days was counted by the annalist. 

In numbers 31 and 32, and two or three items not given in the 
quotations, we have conclusive evidence that 400 days were counted 
to the year by the Annals. They are as follows: 

1 Ah completed the 5th year of the second cycle (25th year) after the 
i revolt (page 171). 

In the following year, ending on 11 Ah, Hunyg died on 12 Carney, 
(ibid. Brinton's translation gives 14 Carney, but this is wrong, as there 
could be no 14 Carney; the original says 12). 

100 clays after was the day 1 Can (page 173). 

20 days later was 10 Queh (page 175). 

The day 11 Imox follows in this year (ibid.). 

The day 9 Batz occurs after this same year (ibid.). 

The year ends on 11 Ah (ibid.). 

As the preceding year ended on 1 Ah, this year began with 2 Yiz, 
and 12 Carney would be the 13th day of the 12th month. One hun- 
dred and twenty daj^s more (or exactly 119) and not 100, as the 
annalist says, would reach 1 Can, the 12th day of the 16th month; 
22 days more would reach 10 Queh, the 14th day of the 17th month. 
The day 11 Imox would be the 8th day of the 18th month, and 9 Batz 
the 18th day of the 20th month, just two days before 11 Ah, the close 
of the year. 


In the year following 5 Ah (number 34), that is to say, the year 
beginning with 6 Yiz, the following events, with dates, are mentioned 
(numbers 35, 36, and 37) : 

On 1 Ganel the Quiches were destroyed. 

On 4 Qnat the chiefs were burned by Tunatiuh (Alvarado). 

On 1 Hunahpu the Spaniards reached Iximche. . 

Five days after. Tunatiuh left the capital; then the Tzutuhils were con- 
quered on 7 Carney. 

Twenty-five days afterward Tunatiuh went forth to Cuzcatan and 
slew Atacat on the day 2 Queh. On 10 Hunahpu he returned, having 
been absent 40 days. 

Iximche was abandoned on 7 Amak. 

Ten days after, on 4 Carney, Tunatiuh began war. 

One hundred and eighty days after the city was abandoned the 29th year 
after the revolt was completed on 2 Ah. 

The day 1 Ganel (or Kanel) was the loth day of the 2d month; 4 
Qnat (or Kat) was the 11th day of the 3d month; 1 Hunahpu the 7th 
day of the 5th month. "Five days after" should be 6 to reach 7 
Carney, the 13th day of the 5th month. "Twenty-five days after- 
wards" (after 7 Carney) should be 21 to reach 2 Queh, 14th day of the 
6th month, and 10 Hunahpu is the 7th day of the 10th month, hence 
the 40 days, if counted from 2 Queh, would be wrong. The 7 Amak 
would be the 3d day of the 12th month, and 4 Carney the 13th day 
of the 12th month. From 7 Amak, the day Iximche was abandoned, 
to 2 Ah, the end of the year (still counting 400 days), was only 177, 
the round number given by the annalist being 180. 

These items of evidence are sufficient to prove, bej^ond any reason- 
able doubt, that the annalist counted 400 days to the year, and that 
the years of the calendar which he used always began with the day 
Yiz. The beginning and ending days of the years would therefore 
be as follows if we start with 12 Yiz, the first year after the Revolt: 




Ending day 


Beginning day 


Ending day 























and so on. 



The next question that arises, and the one of most importance' in 
the discussion, is this: Was the writer justified in indicating thai such 
a calendar as tli is was in use among the Cakchiquels at the coming 
of the Spaniards? On this point we must judge chiefly by internal 
evidence. As what is known in regard to the history of the manu- 
script is given by Brinton in his introduction, it is unnecessary to 

282 MAYAN CALENDAE SYSTEMS [eth. ann. 22 

repeat it here. The writer claims to have been a descendant (grand- 
son) of the ruling chief of the Cakchiquels at the time of the arrival 
of the Spaniards, and was then a youth of probably some 16 or 18 
years. Judging by his method of giving dates, he seems to have been 
familiar with a calendar then in use. Moreover his station would 
indicate that he had been trained in the study of the chronology of 
his tribe. I am, therefore, inclined to accept as substantially cor- 
rect his statements so far as they bear on the calendar system, though 
the traditional portion may be of very little or no historical value. If 
this view be accepted, it may throw some light on one troublesome 
feature of the Maya calendar — the introduction of the multiple 18 in 
counting the months. Why the change from the lunar period to a 
jDeriod of twenty days to the month was made, is not easily accounted 
for, except on the supposition that, having decided for ceremonial or 
other reason to abandon the lunar count, it was natural to follow the 
vigesimal system, hence the 20 days to the month, 20 months to the 
year, and 20 years to the cycle or ahau. The necessity, however, for 
some adjustment between the ceremonial and true year brought about 
at length the adoption of 18 months and 5 added days, and the sub- 
stitution of 18 in place of 20 in time numeration. It seems possible, 
if the annalist be correct in his time count, that the peculiar native 
calendar may have come into use somewhat in this way. 

I can find no grounds whatever for Goodman's assertion that the 
calendar year of the Cakchiquels consisted of 366 days. They may be 
in a historical mention which I have failed to find, but by no possible 
means can this year be made to agree with the calendar of the Annals 
without assuming an arbitrary break in the succession of the daj's at 
the end of each year. 


As I have, in niy paper on the "Mexican and Central American 
Numeral Systems," a brought up the question, How did the Maya 
priests actually perform their calculations relating to time series, some 
of them reaching into millions? I propose to discuss the subject 
somewhat more at length here. As was stated in that paper, these 
calculations sometimes required changing series of days, chuens, 
ahaus, katuns, cycles, and even great cycles (or more correctly units of 
the 1st, 2d, 3d, 4th, 5th, and even the 6th order in the vigesimal sys- 
tem), to years, months, and days, reaching from one given date to 
another. As such calculations could not possibly have been made 
mentally, the authors of the inscriptions and codices must have had 
some method of " ciphering," to use a school-boy term, or of making 
the calculation by marking on some object. As was stated in the 
paper referred to, the only allusion to the subject by an early author- 
ity, so far as is known, is the statement by Landa that they performed 
them u on the ground or some flat thing." 

« Nineteenth Annual Report of the Bureau of American Ethnology. 


As the different kinds of symbols used by the Mayas to express 
numbers have been referred to, I assume that the reader is familiar 
with them. That direct multiplication and division would seem to be 
impossible with their characters where both numbers included units 
above the first order, or, at most, first and second orders, will be 
admitted. The suggestion by Professor McG-ee (referred to in the 
paper on numeral systems) that these operations might have been 
performed by addition and subtraction seems to be the key to the prob- 
lem, as I shall attempt to show. 

That the Mayas could add and subtract numbers expressed in the 
ordinary numeral symbols (dots and short lines) is known from hun- 
dreds of examples in the Dresden codex; and that for these characters 
the} 7 could readily substitute equivalent symbols of other forms in 
use is evident. Take, for illustration, part of a series from plate 
xxiv, Dresden codex (see plate lxxxii), which has been reversed, so 
that it is to be taken from left to right instead of from right to left, as 
in the original. The date below each column is written out, and 
instead of the naught symbol a cipher (0) is inserted: 

(1) (2) (3) (4) (5) . (6) 

• • • • • • 

• • • • • 

• • • • 

• ••• • ••• 

9 Ahau 4 Ahau 12 Ahau 7 Ahau 2 Ahau 10 Ahau 

If we write these in Arabic figures, preserving the relative positions 
and omitting the dates, as those given can be referred to, the series 
will be as follows : 


1 2 










5 (8) 10 


Doubling the first column (8-2-0) we get 16-4-0; adding again 
8-2-0, we get 1-4—6-0; adding again 8-2-0, we get 1-12-8-0 (the 5 in 
this column should be 8, as by adding 8-2-0 to it as thus corrected 
we get 2-0-10-0, the 5th column, etc.). 

If we write the equivalent of each number in days, maintaining the 
same relative positions, and give the sum of each column below (mak- 
ing the correction noted), the result will be as follows: 


7, 200 


14, 400 


5, 760 











2,920 5,840 8,760 11,680 14,600 17,520 

284 MAYAN CALENDAR SYSTEMS [eth. ann. 22 

By adding 2,920 to the first, we obtain the sum of the second column; 
and adding the same number to the sum of the second, we obtain the 
sum of the third, and so on. Ity counting forward 2,920 days from 
9 Ahau, the date under the first column, we reach 4 Ahau, the date 
under the 2nd column, etc. " 

These primary steps are, of course, well understood by readers who 
have given any attention to the subject, but it is necessary to present 
them as leading up to the object in view in the discussion. 

It is evident that • • , or 2,920, is the factor or added number used 

in this series, but the process is carried on by addition. However, 
before we proceed, it is necessary to call to mind certain facts in rela- 
tion to the calendar. The first is that a day of any given name 
returns at every 20th day, whether we count backward or forward, 
but not with the same number; the second, that any given day returns 
with the same day number at every 260th day, whichever way we 
count, but not in the same month nor on the same day of the month 
beyond the first year. As each count reaches Ahau in this instance, 
and 260 is not an even divisor of 2,920, the basal factor must be 20, 
and the day numbers will be different, as we find them to be. Although 
we may not be able always to state why particular factors or counters 
are selected, yet in this case it would seem that 2,920 was chosen 
because this is exactly the number of days in eight years. As the 
dates are therefore just eight years apart, they necessarily fall in years 
having the same dominical day, and, consequently, on the same day of 
the month. However, these specific features must be understood as 
applicable to this particular series, and not as of general application, 
for we shall find series in which there is no reference to the vear; but 
these time periods have a bearing on the practical method used in 
Maya calculations. 

Now, let us see theoretically how, starting with a given date, the 
initial date of a high series may be reached. Nine cycles and the 
lower fractional numbers, counting from 4 Ahau 8 Cumhu as the 
initial date, form the most frequent series of the Copan and Quirigua 
inscriptions. We will try to form such a series, selecting at random 3 
Chicchan 18 Yax, year 1 Lamat, as the terminal date, and 4 Ahau 
8 Cumhu as the initial date. As the former date must be the more 
recent on this supposition, it follows that the count was backward 
(though this is by no means necessary, as it could be forward as well) ; 
so our count in this case will be backward. In order not to make the 
series too long and tedious, we will select as our factor or sum to be 
added — 





This represents a calendar round or cycle of 52 years (18,980 days), 
the given date (3 Chicchan 18 Yax) returning at the end of this 
period. For convenience we make the series ascending toward the 
right, and after a few additions double the columns to make progress 
more rapid. The usual rule is followed; the counter is given as the 
first column; the columns are numbered as a means of reference. 

(1) (2) (3) (4) (5) (6) (7) (8) (9) 

• • • • 

• • • • 

• • • 

0000 00 00 

The counter or first column is added to itself, or doubled, to form 
the second ; the first is added to the second to form the 3rd ; the first 
to the 3rd to form the 1th; but to hasten the i)rocess the} 7 are 
doubled successively from this point to the 8th. As doubling the 8th 
would raise the number above that contemplated, only the number 
necessary to give the 9 cycles is added, but this must be the counter 
(first column) or a multiple of it. The required number is found in 
the 5th column; this added to the 8th gives the 9th. The sum of 
the 9th column, if no mistake has been made, should, counting back 
from 3 Chicchan 18 Yax, bring us again to the same date. 

As a count of a cycle of 52 years (our first column) includes the 
entire series of days and day numbers known to the system, 4 Ahau 
8 Cumhu must be contained therein, and the count to it from the 
date reached must be less than the amount represented by our first 
column. Our next step, therefore, is to ascertain the lapse of time 
from our last date (3 Chicchan 18 Yax) to the next preceding occur- 
rence (as we are counting backward) of 4 Ahau 8 Cumhu. Just what 
method the authors of the inscriptions and codices employed for this 
purpose, as there are more than one, I can not state positively, but 
give one which I am satisfied they could follow. 

They could readily ascertain, as is shown by almost every numeral 
series with a date, that the day 3 Chicchan 18 Yax fell in the year 
1 Lamat, and 4 Ahau 8 Cumhu in the year 8 Ben; hence they could 
easily tell, by counting on their fingers or making marks, that from 
the latter to the former is 18 years and the fractions of the two years — 

• • • • 

the fraction in the former being L98 days or. . .and in the latter 

17 or ' =zr. As the year is represented by q the L8 years would be 

286 MAYAN CALENDAR SYSTEMS [eth. ann. 22 

By adding to this the . . . and r, or together _,we obtain the 

sum represented by _."* Add this to the 9th eolnnm, the 

result is the following number, to wit: '^-- or 9 cycles, 10 katuns, 

14 ahans, 15 chnens, 5 days. 

If no mistake has been made, this number, if we count back from 3 
Chicchan 18 Yax, year 1 Lamat, should bring us to 4 Ahau 8 Cumhu, 
the first day of Goodman's so-called 54th great cycle. Trial proves 
it to be correct, thus : 


9 cycles 1, 296, 000 

1 katuns .' . _ . 72, 000 

14 ahaus > 5,040 

15 chuens 300 

5 days 5 

Total 1, 373, 34? 

Subtract 72 calendar rounds 1 , 366, 560 

Remainder : 6, 785 

Counting back this number of days from 3 Chicchan 18 Yax 
brings us to 4 Ahau 8 Cumhu. Turning to Goodman's ''Chronolog- 
ical Calendar," 54th great cycle, 9th cycle, 10th katun, and 14th 
ahau, we find the date is 10 Ahau 18 Mac. Fifteen months and 5 
days from this just reaches 3 Chicchan 18 Yax. The series is there- 
fore a correct one, formed upon the same plan as those of the Dresden 
codex, and without using anything not in the reach and comprehen- 
sion of the aboriginal artist. 

The series on plate xxiv of the Dresden codex (our plate lxxxii) 
appears to close with a minor addition (in the lower left-hand corner) 
to reach the desired date, just as the theoretic one given above, ex- 
cept that in this case the count is forward. The series includes the 
right half of the plate, and reads from right to left and by lines from 
the bottom upward, closing with the lines in the lower left-hand 
corner. Here the steps have been in part from 1 Ahau to 1 Ahau, 
hence with 260 as the primal factor. The last column is 9-9-10-0-0, 
then follows the number 6-2-0, 4 Ahau 8 Cumhu. The latter number 
changed into days is the lapse of time from 1 Ahau 18 Kayab, the 
last preceding date, to 4 Aha'u 8 Cumhu. However, as there are some 
unusual features in regard to the additions in a part of this series, at- 
tention will asrain be called to it a little farther on. 


In order to show that resort was had to increasing the added num- 
ber to shorten the process, as was done in the theoretic example, the 
following example is given from plates lxx and lxxi of the Dresden 
codex. Ordinary numerals are used in place of the symbols, and the 
series, which in the codex ascends from right to left, is reversed ; the 
days below the columns are also given : 























18 ■ 

9 Oc 9 Eb 9 Ix 9 Ix 9 Ix 9 Ix 9 Ix 

It will be seen by subtracting that the difference between the first 
and second columns and between the second and third is 1-17-2, or 
1 ahau, 17 chuens, 2 days, equal to 702 days, while the difference 
between the 3d and 4th columns is 2-18-9-0, or 2 katuns, 18 ahaus, 
9 chuens, days — equal to 21,060 days; and that the difference 
between the 4th and 5th, the 5th and 6th, and the 6th and 7th is, in 
each case, 1-19-0-0, or 14,040 days. There is therefore an increase 
of the added number or factor in passing from the 3d to the 4th 

It will be noticed that the days below the 1st, 2d, and 3d columns 
differ, while from this point onward they are all 9 Ix. The change in 
this respect requires a change in the counter. Why the counter was 
made larger in passing from the 3d to the 4th column than between 
the remaining columns is not clear, as the difference between the 
3d and 4th columns would have reached the desired day, 9 Ix. It is 
possible that the month date, though it does not appear, was here 
taken into consideration. Assuming that the first 9 Ix (under the 
3d column) was 9 Ix 2 Pop, year 8 Ben, the count forward of 1-19-0-0 
would reach 9 Ix 12 Chen in the year 7 Akbal, while the count for- 
ward of 2-18-9-0 would reach 9 Ix 17 Mac, year 13 Ezanab. As the 
first counter (702) is not a multiple of 260 or of 20, it must have been 
based on 13, one of the factors of 260. The counters 14040 and 
21060 are multiples of 260; and there is possibly something in the 
fact that the former (14040) is 54 times 260 and that the first counter 
(702) is 54 times 13. Although Ave are not able at present to solve all 
these problems, it is evident that the author of the codex increased 
the counter as he proceeded, presumably to shorten the process. 

The series appears to close with two columns in the upper middle 
portion of plate lxx, the dates here having the month given. With 
these (notwithstanding the obliterated portion of the scries) wo might 
determine the true dates of the portion given above, and thus possibly 
solve, to some extent, the problems mentioned; but unfortunately 
there are so many errors in these two columns that it seems impossi- 
ble to determine the true numbers and dates. Thechuen number, or 



[ETH. ANN. 22 

number in the place of the second order of units, is 18 or 19 (there 
being a space where one dot may have been obliterated), either of 
which is wrong. The date below one is 9 Ix 20 Pop, the other is 9 Ix 
13 Pax or Tzec, both of which are wrong, as Ix is never the 13th or 
20th of the month. 

A good example of this method of increasing the counter as the 
series proceeds is found on plates lxx-lxxiii of the same codex. 
Although this runs from right to left, we give it here in reverse order 
and in ordinary figures as follows : 














19 [1?] 


4 Caban 

4 Ik 

4 Manik 


4 Caban 

4 Ik 

















4 Manik 


4 Caban 

4 Ik 

4 Manik 










11 [12? 








4 Caban 

4 Ik 

4 Manik 


4 Caban 

00 CO 

T— 1 














4 Ik 

4 Manik 


4 Caban 

4 Ik 















4 Manik 


4 Caban 

4 Ik 

4 Manik 










































The figures in parenthesis are merely arbitrary numbers given to 
the columns as a means of reference. The counter is 3-5, or 65 days, 


from the first to the 28th column; but here a change takes place; the 
amount at this point, being 5-1-0, or 1,820 da3~s, is doubled to form the 
29th column, and is again added to form the 30th. Here again occurs 
an increase in the counter, in this case a large one, viz, to 1-5-5-0, or 
9,100 da} T s; but at the next step the added number to form the 32nd 
column is only 1-0-4-0, or 7,280 days, just one-half of the 31st col- 
umn. This counter is used to the end of the series; however, the 8 in 
the 36th column is an evident mistake; it should be 7. 

The number 65 is a very common counter in this and other codices; 
in this case 13 is the basal factor. In the other counters 260 is the 
permanent factor. The first counter, which is just one-fourth of the 
second, always reaches a day with the same number, though not the 
same day — but repeating by series of four. However, aside from these 
questions, we have the fact of the increase of the counter in the proc- 
ess, to show which was the object of calling attention to the series. 

Returning now to the series on plate xxiv (our plate lxxii), to 
which reference has been made, I call attention to the unusual changes 
in the counter or added number. The series in the fourth tier from 
the bottom, given in the way adopted above, is as follows: 


















1 Ahau 

1 Ahau 

1 Ahau 

The values of the different units and sums of the columns are as 

follows : 

144, 000 

7,200 28,800 64,800 36,000 

1,800 4,320 3,960 5,040 

100 160 140 80 

9,100 33,280 68,900 185,120 

It will be found by trial that the greatest common divisor of these 
totals is 260, and that it is contained in the first total 35 times; in the 
second, 126; in the third, 265, and the fourth, 712 times. Although 
each step must have required long and tedious additions — no two 
having a common added number or multiple thereof — and the reason 
for thus varying the added number is not apparent, yet it is evident 
that the aboriginal scribe chose 260 as the factor to be used, and also 
that the desired result could be reached by successive additions. In 
fact, the series and the others we have noticed seem to be mere records 
of the steps in the process of determining the lapse of time between 
two widely separated dates. 

These examples are sufficient to show that all the series in the 
codices and inscriptions could have been formed by the aboriginal 

22 etip— 04 11) 


authors with their numeral symbols by addition and subtraction. It 
may also be added that the evidence presented to show this is fitted 
to impress us with the belief that some, if not all, of the series of the 
Dresden codex are but records of the process of calculation. 

There arises in connection with this examination a question, the 
proper answer to which may, if determined in accordance with a view 
that has been expressed, have an important bearing on the history of 
the Mayan tribes. 

On several of the plates of the Dresden codex there are what appear 
to be supplemental series connecting with the so-called " normal date," 
4 Ahau 8 Cumhu. However, the discussion of this question will come 
more appropriately under the next section, which relates to the signifi- 
cation of the series of the inscriptions. 


Why were these series formed? What is their signification? These 
are questions impossible for us to answer satisfactorily with our pres- 
ent knowledge of the subject. It may be possible, however, to limit 
the inquiry by certain considerations. 

Our first question is, Were they intended, by the initial or terminal 
days, to refer to actual dates bearing some relation to events in the 
history of the respective tribes to which they pertain? By the term 
" initial dates" I allude to the dates from which the series (whether 
initial or minor) were counted, and by "terminal dates" to those 
which follow the series in counting forward. The latter are assumed 
to be later in actual time than the former. 

That the initial date may be thrown back any desired distance in 
time, is admitted, as for example, we may take as our initial date the 
beginning of the Christian era (A. D.), or the supposed initial date of 
the world era (A. M.), or any other beginning date which, through 
fancy, tradition, or mythology, has been adopted or arbitrarily chosen 
by different peoples. It is not necessary, therefore, that we should 
assume that the initial dates of the Mayan codices and inscriptions 
have any reference to historical or even supposed historical. events. 
That such an assumption would be preposterous is shown by the fact 
that several of these dates reach back in time 33,000 years, and a 
large proportion of those of the inscriptions nearly 4,000 years, and 
others to a still more distant time. The initial dates must therefore 
relate, as will be conceded, to some assumed date, traditional or myth- 
ological, or arbitrarily chosen, according to the fancy of the calcu- 

Do the terminal dates refer to events or incidents in the history of 
the tribes — events which were noted down by the scribes sufficiently 
near the time of occurrence to give the proper or probable dates 


If we take the terminal dates of the initial series at Quirigua (omit- 
ting from consideration those of the minor series) we find the differ- 
ence between the earliest and latest, with two exceptions to be 
noticed, is only some 83 or 84 years. This difference is so moderate 
as to be entirely consistent with the idea that the dates were engraved 
near the time of the events or incidents to which they refer, if, in fact, 
this was the object in giving them. The two excepted are numbers 6 
and 10 of the list given below. The calculation I give is based on 
what seem to be the reliable series and dates, leaving out of consid- 
eration the exceptional and doubtful series. Comparing the earliest 
and latest of those at Copan, we find the difference to be about 
222 years. This is by no means extravagant, hence the dates may 
refer to historical events. When we come to those at Palenque, we 
find the difference — even excluding the most recent date, which Good- 
man admits is doubtful — to be over 3,800 years. Although a differ- 
ence in dates as great or greater than this has been found in the 
inscriptions of the ruins of Egypt and Assyria and accepted as reason- 
ably correct, no archeologist of the present day not carried away by 
some extravagant theory will believe that inscriptions were chiseled 
at Palenque at dates 3,800 years apart in actual time, the earliest 
(counting from the coming of the Spaniards) going back more than 
2, 200 years before the Christian era. 

Now, it is the opinion of Goodman and Seler that the terminal dates 
of the inscriptions (the latter excepts those at Palenque, as explained 
below) refer to the times when the monuments were erected or the 
inscriptions chiseled. The assertion of the former on this point (pages 
147-8) is as follows: 

Particular emphasis is intended to be laid upon ' ' initial ' ' dates in the foregoing 
estimate. There are two kinds of dates in the Archaic inscriptions. The dates 
of one character, and those of most frequent occurrence, appear in the body of the 
texts, and designate the points from or to which the reckonings extend. Some- 
times they are but a day apart; at others, they are a few months or years, while 
occasionally a flight is made over thousands of years and back again, with the 
ease and swiftness with which in Eastern story the couch of the prince is trans- 
ported by genii. These dates have no significance beyond their relation to other 
dates and the corresponding reckonings. 

But with the other class, the initial dates, as Maudslay has very appropriately 
named them, it is quite different. The inscription on nearly every temple, stela, 
and altar begins with one of them, reciting the great cycle, cycle, katun, ahau, 
chuen, month, and day. Such conspicuousness and circumstantiality, in my esti- 
mation , could have but a single purpose — that of recording the date at which the 
monument was erected. Some of the stelae have different initial dates on oppo- 
site sides, but in these instances one date is reckoned from the other, the later one 
undoubtedly designating the time of dedication. I think there is nothing we can 
assume with more assurance of certainty than that these initial series mark the 
date of erection of the respective monuments. 

Taking this for granted, also, we will turn to the inscriptions and see to what 
these conclusions lead. The latest initial date is found on a stela at Quirigua. 
It is oo-ii-lO-S-lBX^O— 7 Ahau-18 Pop. That is 2.X40 years subsequent to the 

292 MAYAN CALENDAR SYSTEMS [eth. ann. 22 

average of initial dates in the other Quirigua inscriptions. The next latest 
initial date is on a restored stairway in one of the temples of Palenque. It is 
55-3-18-12-15 X 12—8 Eb-15 Pop. That is 7,082 years later than the earliest initial 
dates at Palenque. These are long periods, but the limit is not yet reached. In 
the museum at Leyden is the misnamed " Yucatec " stone, exhumed in digging a 
cut on the line between British Honduras and Guatemala, about a hundred miles 
from Copan. It is a slim slab of jadite, about a foot long and four inches wide, 
if my recollection of it is correct. Both sides are inscribed in rather a rude man- 
ner, the rudeness apparently being more attributable to the hardness of the stone 
than to a lack of skill in the artist. The carving on the front represents a warrior 
trampling an enemy under his feet. The stone, therefore, is evidently a memo- 
rial of some victory or conquest. The inscription on the back consists of an ini- 
tial date in the Archaic form and characters. It is 53-8-14-3-1x12 — 1 Eb-5 Zac. 
That is 8,383 years anterior to the latest initial date in Quirigua. Now, if in 
accordance with my theory respecting the era of the Archaic cities the 2,348 years 
that have elapsed since that Quirigua date was made be added to the above period, 
we shall arrive at the time when that ancient Maya conqueror trod his enemies 
under foot — 10,731 years ago — the oldest historical date in the world. 

Dr Seler's opinion on this point is expressed in the following quo- 
tation from his paper in the Zeitschrift fur Ethnologie, Heft 6, 1899 : 

I have, in conclusion, now to speak of the relation in which the various monuments 
which we have become acquainted with stand to each other. Here at the outset is 
to be kept in mind the noteworthy difference which exists between the altar plates of 
Palenque and the remaining monuments. I have already mentioned that the initial 
series of all monuments which we are able to read contain in the first member 
the multiplier nine; and I can add that the same holds also for the stela? of 
Quirigua (which I have not yet been able to treat of, as they have not yet been pub- 
lished in Maudslay's work) and for stela 6 of Copan, excavated by the engi- 
neers of the Peabody Museum. On the altar plates of Palenque, on the contrary, 
so far as we have been able to decipher them, there stands in the first member 
the multiplier one. If, as indeed is a priori most probable, the date designated 
at the> end of the first series gives the time of erection of the monument in ques- 
tion, then we must conclude that all other monuments within the tenth cycle 
after the beginning and normal date 4 Ahau 8 Cumku — the Temple of the Cross 
II of Palenque, the Temple of the Sun, and perhaps also the Temple of the Cross 
I — were constructed within the second cycle after the beginning and normal date. 
In other words, we must conclude that between the time of the erection of the 
temples of Palenque and of all the other monuments there lies a period of 
about 3,160 years; that the temples of Palenque are about 3,160 years older 
than the monuments of Copan and Quirigua, and than the steps of the tow- 
ering palace of Palenque not far from the temple. This is, in itself, not probable, 
and all the less so as one would, from the style of the hieroglyphs and figures, be 
rather inclined to explain the temples of Palenque as younger than the stela? of 
Copan. The solution of the riddle may be a different one. It may be that, in the 
initial series of the temples of Palenque, the end date does not represent the date 
of erection of the temple, but an earlier sacred date which it had been determined 
to bring into view. It may, however, also be that the time of the erection of the 
monument was brought into view, not through notation of the actual traditionally 
accepted distance from the normal date, but as it were in arithmetical fashion 
through notation of one difference which led from the normal date to a day of 
this name. 

The end dates of all the remaining monuments which we are able to read fall, 
as said, within the tenth cycle after the beginning and normal date 4 Ahau, 8 




It is apparent from these quotations that both Goodman and Seler 
hold the opinion that the terminal date in an initial series is intended 
to indicate the time when the monument was erected, though the 
lapse of time given by Goodman (who does not seem to object to long- 
periods) to the dates of erection of the various monuments differs 
very widely from that allowed by Seler. The differences I have 
indicated are, as was stated above, limited to those which remain 
after rejecting those which seem doubtful. 

Let us discuss this question on the data furnished by the inscrip- 
tions and Dresden codex, taking, where there are not good grounds 
for objecting to them, the interpretations of the initial series by Good- 
man and Seler. Differences in the numbers of the periods or orders 
of units below that which Goodman terms " katun" have no bearing 
in this discussion. In order that the reader may have the data before 
his eye, I give below a list of the initial series, retaining, for con- 
venience, Goodman's great cycle numbers. The numbers at the left 
are merely for reference. 


(1) 53-12-19-13- 

4- 0, 

8 Ahau 

18 Tzec. 

Temple of the Cross. 

(2) 54- 1-18- 5- 

3- 6, 

13 Cimi 

19 Ceh. 

Temple of the Sun. 

(3) 54- 1-18- 5- 

4- 0, 

1 Ahau 

13 Mac. 

Temple of the Foliated Cross 

(4) 54- 9- 4- 0- 

0- 0, 

13 Ahau 

18 Yax. 

Temple of Inscriptions. 

(5) 55- 3-18-12- 



15 Pop. 

Inscribed steps of palace. 


(6) 54- 9- 1- 0- 

0- 0, 

6 Ahau 

13 Yaxkin. 

Stela C, west side. 

(7) 54- 9-14-13- 


12 Caban 

5 Kayab. 

Stela F, west side. 

(8) 54- 9-14-13- 


12 Caban 

5 Kayab. 

Stela E, west side. 

(9) 54_ 9_i6- 5- 

0- 0, 

8 Ahau 

8 Zotz. 

Stela J, back. 

(10) 54- 9-16-10- 

0- 0, 

1 Ahau 

3 Zip. 

Stela F, east side. 

(11) 54- 9-16-13- 


8 Caban 

5 Yaxkin. 

Stela D, west side. 

(12) 54- 9-17-10- 

0- 0, 

12 Ahau 

8 Pax. 

Animal B. 

(13) 54- 9-17- 5- 

0- 0, 

6 Ahau 

13 Kayab. 

Stela A, east side. 

(14) 54- 9-17-15- 

0- 0, 

5 Ahau 

3 Muan. 

Animal G. 

(15) 54- 9-18-15- 

0- 0, 

3 Ahau 

3 Yax. 

Stela K. 

(16) 54-13- 0- 0- 

0- 0, 

4 Ahau 

8 Cumhu. 

Stela C. 

(17) 54- 9-16-15- 

0- 0. 

7 Ahau 

18 Pop. 

Stela D, east side. 


(18) 54- 9- 5- 5- 

0- 0, 

4 Ahau 

13 Zotz. 

Stela D. 

(19) 54- 9- 9-10- 

0- 0, 

2 Ahau 

13 Pop. 

Stela P. 

(20) 54- 9-12- 3- 

14- 0, 

5 Ahau 


Stela I. 

(21) 54- 9-12-12- 

0- 0, 

1 Ahau 

8 Zotz. 

Stela J, west side. 

(22) 54- 9-12-16- 

7- 8, 

3 Lainat 

16 Yax. 

Altar K. 

(23) 54- 9-13-10- 

0- 0, 

(7 Ahau 

13 Cumhu.) 

Stela J, east side. 

(24) 54- 9-14-10- 

0- 0, 

5 Ahau 

3 Mac. 

Stela F. 

(25) 54- 9-14-19- 

8- 0, 

12 Ahau 

is Cumhu. 

Stela A. 

(26) 54- 9-15- 0- 

0- 0, 

4 Ahau 

13 Yax. 

Stela B. 

(27) 54- 9-15- 0- 

0- 0, 

4 Ahau 

13 Yax. 

Altar S. 

(28) 54- 9-16-10- 

0- 0, 

1 Ahau 

8 Zip. 

Stela N. 

(29) 54- 9-16- 5- 

0- 0, 

8 Ahau 

8 Zotz. 

Stela M. 

(30) 55- 2- 6- 0- 




Altar L. 

(31) 55-13- 2-18- 

0- 0, 

6 Ahau 

18 Kayab. 

Stela C. 

2V)4 MAYAN CALENDAR SYSTEMS [eth. ann. 22 

Tlie Ley den Stone 

(32) 53- 8-14- 3- 1-12, 1 Eb 5 ?(Yaxkin?). 

Goodman also mentions (p. 148) the following as at Quirigua: 

(33) 55- 3-19- 2- 0- 0, 7 Ahau 18 Pop. Stela ? 

Examining this list, we see that the terminal dates of 24 out of the 
33 series fall in the 10th (Goodman's 9th) cycle from 4 Ahau 8 Cumhu, 
the initial day of Goodman's 54th great cycle. It can not be doubted, 
therefore, as we also find the same initial date the most prominent 
one in the Dresden codex, that, for some reason unknown to us, it 
was selected by the people who made the inscriptions and codex as 
their principal era date. As the 24 series ending in the 10th cycle run 
back from the earliest terminal date (number 6) 9-1-0-0-0, or 3,570 
years, and from the latest terminal date (number 15) 9-18-15-0-0, or 
3,920 years, it is evident, as has been stated above, that the normal 
date (4 Ahau 8 Cumhu) selected as the commencement of this era could 
have no reference to an historical event remembered by the Mayan 
people. Even if we suppose that the last of these inscriptions was not 
chiseled until the close of the fifteenth century, this would carry back 
the era date 2,400 years before the Christian era. The only safe and 
reasonable conclusion, therefore, is that the initial date was arbitrarily 
selected for some mythological, mystical, or arithmetical reason. It 
is especially worthy of notice, however, that the lapse of time between 
the terminal dates of the earliest and latest of these series is only 
about 350 years, and, if number 6 be omitted, less than 90 years. This 
fact would seem to give color to the suggestion of Goodman and Seler 
that the terminal dates of the initial series refer to the time the monu- 
ments were erected. Nevertheless, there are some serious difficulties 
to be overcome before this theory can be considered as satisfactorily 
established, some of which it will be my object now to point out. 

So far as the foregoing list is concerned, all the series which begin 
with 9 cycles (the 54 indicating the so-called great cycle is omitted 
from consideration) have, beyond question, the initial date 4 Ahau 
8 Cumhu. It must be remembered, however, that this date returns 
at the end of every count of 18,980 days, or 52 years. Now, the ques- 
tion arises (and it is a crucial one in this discussion), Does the count 
in each one of these series go back to identically the same 4 Ahau 8 
Cumhu, or merely to any 4 Ahau 8 Cumhu? If, as I think I have suc- 
cessfully shown, the so-called ahaus, katuns, cycles, and great cycles 
are not absolute time periods, recognized as such in any Mayan time 
system, but are mere orders of units in the Mayan method of numera- 
tion, these counts would be precisely like the following in our ordinary 
time system: Thursday the 15th day of the 7th month of the 48th year 
of the century. What century? Or 1,025 years, 7 months and 15 days 
from December 25th to Thursday the 9th da} r of the 8th month. It 
is evident that without the first or last date being fixed in some recog- 


nized calendar the 1,025 years, etc., may be pushed backward or for- 
ward at will. Hence a Mayan scribe may write 9-15-0-0-0 from 4 
Ahau 8 Cumhu to 4 Ahau 13 Yax (as in number 2G); and 52 years later 
another may write the same series, as in number 27, and both will 
be strictly correct, but the 4 Ahau 8 Cumhu of the first will be 52 
years earlier than that of the second. The mere fact, therefore, that 
4 Ahau 8 Cumhu is reached by counting back the different numeral 
series is not evidence that in each case identically the same 4 Ahau 
8 Cumhu is reached. Other evidence having some bearing on the 
question must be introduced to establish this identity. The only fact 
apparent in the series themselves which seems to favor the theory of 
identity is that each runs back 9 cycles plus the minor numbers. 
This undoubtedly favors the theory of identical date. 

Let us turn now to the Dresden codex, and give attention to what 
I have termed subsidiary series; that is to say, short series ap- 
parently, as was suggested in the theoretical series given above, 
intended to connect with 4 Ahau 8 Cumhu. As I have expressed 
doubts as to the correctness of Seler's suggestion about that on plate 
lxix, attention is called to the long compound series on plate xxiv 
(see our plate lxxxii). This series begins at the right-hand edge of 
the bottom section and runs leftward to the middle; it then passes to 
the next section above, and so on to the top of the page, the conclud- 
ing column being that in the lower division of the left-hand portion. 
No months are given except at the bottom of the long number col- 
umns and the one short column in the lower left-hand portion of the 
plate. The last date standing in the lower left-hand corner is 4 Ahau 
8 Cumhu, and over it is the number series 6-2-0 (the symbol in a 
red loop). The next date to the right is 1 Ahau 18 Kayab; this 
stands under the numeral series 9-9-16-0-0. Counting back from 
4 Ahau 8 Cumhu, the short series, 6 ahaus, 2 chuens, days, or 2,200 
days, we are brought to 1 Ahau 18 Ka} 7 ab, while if we count forward 
from the same date it brings us to 7 Ahau 18 Cumhu, which shows 
the backward count to be the correct one, if the design of the artist 
was to connect the two series; moreover, the count of the long series, 
if made toward the right, is backward. 

We know that in all the series given in the above list, where 4 
Ahau 8 Cumhu is the principal date, it is the initial day and the 
numeral series follows it; in other words, the count must be backward 
to reach it. Taking number 15 of the list — Stela K of Quirigua — 
54-9-18-15-0-0 — we find that the terminal date lies 3,920 years subse- 
quent to 4 Ahau 8 Cumhu. Turning to the last column of the series 
on plate xxiv of the Dresden codex, which is 9-9-16-0-0, we find that 
the count, when the short series of 2,200 days is added, reaches back- 
ward from 4 Ahau 8 Cumhu 3,750 years. In other words, we count 
forward in the codex 3,750 years to 4 Ahau 8 Cumhu, and in t lie 
inscription series forward from this date 3,920 years, making the total 

296 MAYAN CALENDAR SYSTEMS [eth. ann. 22 

lapse of time from the beginning date of one and the ending date of 
the other 7,670 years. Is it at all probable that the one 4 Ahan 8 
Cumhn is the same in actual time as the other? That the count is 
necessarily forward in the codex series may be proved thus: The 
last column (that in the lower left-hand portion) reaches back to the 
initial date, which is found to be 1 Ahau 18 Kayab, the same as the 
terminal date which stands below the column. Now if the supposi- 
tion be correct that, as is usual in this codex, this column is the sum 
of the series, and there is no mistake on the part of the aboriginal 
artist, the first number column, that in the extreme lower right-hand 
corner of the plate, 8-2-0, 9 Ahau (the symbol appears to be 8, but 
the fourth dot is hid by the red border line, as can easity be shown by 
the steps from date to date toward the left) should give the exact lapse 
of time from 1 Ahau 18 Kayab. Counting forward 8-2-0, or 2,920 
days, from 1 Ahau 18 Kayab, year 2 Akbal, we reach 9 Ahau 18 Kayab, 
year 10 Akbal, the date under this first column. Counting forward 
2,920 days (the difference between the first column and the next one 
to the left) from the last date (9 Ahau 18 Kayab), we reach 4 Ahau 18 
Kayab, year 5 Akbal, the date under the second column. Counting 
back the sum of this second column — 5,810 days — we reach, as we 
should, 1 Ahau 18 Kayab, the initial date. 

As further proof that the series is continuous and the count for- 
ward, let us select at random the third column, counting from the 
right, of the third section from the bottom, to wit, 4-8-1-0, 11 Ahau. 
Counting forward 32,120 days, the sum of this column, from 1 Ahau 
18 Kayab, we reach 11 Ahau 18 Kayab, year 12 Akbal — the day under 
this column. If we take the column immediately above (third from 
the right in the fourth division from the bottom of the page) which 
reads 9-11-7-0, 1 Ahau, equal to 68,900 days, and count forward from 
the initial date 1 Ahau 18 Kayab, we reach 1 Ahau 13 Mac, year 9 
Lamat. Subtracting this column from that to the left of it — 

1_5_14_ 4_o 
9_H_ 7_o 

16- 2-15-0 

we find the remainder to be 16-2-15-0, or 116,220 days. Counting for- 
ward this number of days from 1 Ahau 13 Mac, the date under the 
third column from the right, we reach 1 Ahau 18 Uo, year 3 Akbal, 
the date under the last or fourth column from the right, which proves 
the steps thus far taken to be correct. 

Although the upper division is too nearly obliterated for any of its 
columns to be used to calculate forward to the final column, we can 
do this as correctly by subtracting the last column of the fourth 
division from the terminal column of the entire series, thus — 

9-9-16- 0-0 
1_5_14_ 4_o 

8-4- 1-14-0 




• • •• 

Using this remainder, which amounts to 1,181,440 days, subtracting 
from it 62 calendar rounds or 1,176,760 days, which leaves a balance 
of 4,680 days, and counting forward from 1 Ahau 18 Uo (the date 
under the left column of the fourth division), we reach 1 Ahau 18 
Kayab, the date under the final column in the lower left portion of 
the plate. No doubt, therefore, is left that the count in this long 
series is toward the left and forward in time, and that the 1 Ahau 18 
Kayab under the final column is 3,744 years later in time than the 
initial date, which is also 1 Ahau 18 Kayab. 

Counting forward from this terminal date the short series in the 
extreme lower left-hand column (2,200 days), we reach 4 Ahau 8 
Cumhu, the date in the corner below this short column. It is certain, 
therefore, that 4 Ahau 8 Cumhu is the terminal date of the long 
series on this plate. Is it the ' ' normal date," the same 
initial 4 Ahau 8 Cumhu from which the series of in- 
scriptions are counted? To show that Goodman's cal- 
culations agree exactly with this result, we have only 
to count back on his chronological tables from 4 Ahau 
8 Cumhu, the first day of his 54th great cycle, the 9 
cycles, 9 katuns and 16 ahaus of the final large col- 
umn and the 6 ahaus of the short column. This will 
reach 2 Ahau 13 Pop, the first day of the 18th ahau 
of the 9th katun of the 3rd cycle of his 53rd great 
cycle. Counting back from this the two, months of 
the short column we reach 1 Ahau 18 Kayab, the 
initial day of the long series of the codex plate. 

This fact will tend to throw a strong doubt on the 
theory of Goodman and Seler in regard to the signifi- 
cation of the series. Moreover, if we turn to plate 
lxx of the codex we see high numbers, some reach- 
ing to 8 and others to 9 cycles, one being as high as 
9-19-11-13-0. These are followed bv a short sub- 
sidiary series ending with 4 Ahau 8 Cumhu. Here, 
then, this "normal date" comes after the long series of 3,937 years, 
and if Seler's idea that the 4 Ahau 8 Cumhu in the texts of plates lxi 
and lxix is to be connected with the high series in the serpent 
figure be correct, then it must stand at the commencement of a period 
extending back from the terminal date some 33,900 years. 

As an example clearly illustrating the statements in the preceding 
paragraph occurs on plate xliii of the Dresden codex, I shall notice 
it here before passing from the point under discussion. This consists 
of a single column shown in figure 166. At the head of the column is 
the day 3 Lamat; immediately below is a figure with a turned-up 
nose, probably a conventionalized tapir head, which, as it occupies 
the same relative position as the great cycle symbol in the inscrip- 

Fig. 10(5. Column 
from plate xliii, 
Dresden codex. 



[ETH. ANN. 22 

tion, may, and in fact probably does, stand for the same purpose here. 
Following the latter, reading downward, is the series 9-19-8-15-0 
(9 cycles, 19 katuns, 8 ahaus, 15 clmens, days); next comes the day 
3 Lamat, which is followed by the short series 17-12 (17 clmens, 12 
days), the column ending with the day 4 Ahan, though no month 
symbol is given. 

Assuming the date at the bottom to be 4 Ahau 8 Cumhu, we count 
back 17 months and 12 days(=352 days) from this date. This brings 
us to 3 Lamat 1 Uayeb in the year 7 Lamat. Counting back from 
the latter date 9-19-8-15-0, or 1,435,980 days, we reach 3 Lamat 
11 Muan, year 12 Ezanab, the day standing at the head of the column. 
It is true that we have no absolute proof that the terminal date (4 
Ahau) is intended for 4 Ahau 8 Cumhu, as the count will give the 
same result from any other 4 Ahau. The column given is the sum of — 
that is to say, includes — the long series which occupies the right por- 
tion of the middle section of plate xliii and the left portion of the 
middle section of plate xliv, and seems to be here precisely what an 
initial series is in the inscriptions. This supposition, which seems to 
be confirmed by the tapir-head symbol, which apparently stands for 
the great cycle, is in direct opposition to the assumption that 

Fig. 167. Centerpieces of great cycle symbols. 

the terminal 4 Ahau is the initial day of a great cycle. On the other 
hand, the assumption that it is the initial day of a great cycle, as 
Seler seems to think, necessitates the conclusion that the date 3 
Lamat 11 Muan, from which the count of the series starts, is not the 
beginning of a great cycle, or that great cycles may overlap one 
another. The latter conclusion would indicate that the starting point 
is arbitrary, and that the supposed time-periods are simply orders of 
units in expressing numbers. 

At any rate, if the 4 Ahau is assumed to be 4 Ahau 8 Cumhu, the 
whole of the series lies back of, or anterior to, the commencement of 
Goodman's 54th great cycle. 

As an indication that the conventional tapir head on plate xliii of 
the Dresden codex is used as a great cycle symbol, attention is called 
to the centerpieces of the three great cycle symbols shown in figure 
167, the one marked a being from the east side of Stela F, Quirigua; 
b from Stela N, Copan, and c from Stela 6, Copan. The resemblance 
to the codex symbol is too strong to be overlooked. 

In addition to these facts which seem to stand against, or at least to 
render doubtful, the supposition that 4 Ahau 8 Cumhu, when stand- 


ing as the initial or terminal clay of a series, is to be taken as the 
date of the chosen era, there is the additional fact that in quite a 
number of the inscriptions there are series connected with, but sub- 
sequent to, the initial series, sometimes running into the hundreds of 
years. If the terminal date of the initial series designates the date of 
erection, then the other subsequent dates must have been chiseled 
after the monument was erected. This would require the supposition 
that the tablets at Palenque were quarried and dressed to a particu- 
lar size with a profound knowledge of or keen foresight as to the 
additional space that would be needed in the coming years. 

Such are some of the difficulties that stand in the way of the theory 
advanced by Goodman and Seler as to the signification of the inscrip- 
tions. Nor are these all the difficulties ; others appear when we discard 
Goodman's theory of a great chronological system and look upon his 
so-called time-periods as but orders of units, and count, as should be 
done, 20 of the 5th order of units (cycles) to one of the 6th order (great 
cycle). However, notwithstanding these serious difficulties, the 
theorj^, if a little more generalized, so as to apply to the latest date 
in the inscription as that denoting the time of erection or event com- 
memorated, is perhaps the most acceptable which has been pre- 
sented, though it be very doubtful. Many of the long series in the 
Dresden codex appear, in fact, to be records of the steps of calcula- 
tion in finding the lapse of time between widely separated dates, 
seemingly for amusement or mystical purposes. The author of the 
Dresden codex seems to have been of a mathematical turn — far more 
so than the authors of the Troano and Cortesian codices, which fact 
probably accounts for the long series in the former; and it may be 
added that a strong mathematical turn of mind has iirobably led Mr. 
Goodman to form his grand but, unfortunately, imaginary Mayan 
chronological system. 


Attention is called again to figure 157 (page 253), showing an 
inscription found at Xcalumkin, Yucatan, by Mahler and photo- 
graphed by him. A copy was obtained by Dr Le Plongeon and pub- 
lished in his "Queen Moo" (page 80, plate xxv), but without any 
particulars or attempt at explanation. As Mahler has not, so far 
as I am aware, published any account of this discovery, and I am 
indebted to Dr Saville for the copy used, I can only refer to the 
inscription, which is certainly interesting in several respects. 

It is apparent at a glance that the majority of the symbols differ 
very considerably from those at Palenque, Tikal, Copan, and Quirigua 
to which reference has been made in my previous paper and the first 
part of this paper. So great is the difference that we are unable to 
say whether the first symbols, Al to B2, are numeral characters repre- 
senting an initial series. That the part of A3 which is a cartouch 
inclosing a serpent is to be taken as a day symbol may be safely 



[ETH. ANN. 22 

assumed. If this surmise be correct, it is a type different from any 
hitherto found in a Mayan inscription. If a Mayan day symbol, it 
must, bej^ond any reasonable doubt, represent Chicchan, which is the 
only day in the calendar that has received the interpretation "Serpent," 
and is that which corresponds in position with Cohuatl in the Mexican 
calendar. If this conclusion be correct, it confirms Brinton's inter- 
pretation of the name "Chicchan" (Native Calendar, page 25). 

The important glyphs of this inscription are the two at the bottom, 
A 6 and B6. These I think may safely be read "8 Caban 4 Zotz," 
and in this I am glad to say that Saville agrees with me. Whether 
the determination of the month symbol be correct or not, the four 
dots over it are clear and distinct, showing the day to be the 4th of 
the month. There can scarcely be any doubt that the day symbol is 
that of Caban, which can only be the 4th day of the month in years 

beginning with Ix. This conforms to the 
calendar of the Troano and Cortesian codices 
and that used by Landa, in which the domin- 
ical days are Kan, Muluc, Ix, Cauac. 

This is a very important fact, which, if 
corroborated by other discoveries, will carry 
back the use of the Yucatec calendar to an 
earty date. I was inclined to the opinion 
that this calendar was of comparatively 
recent date, but this evidence, if accepted, 
must carry it back to the era of the inscrip- 
tions, and place it, in time, parallel with that 
of the other sections. 

A single date, it is true, is slender evi- 
dence on which to base a conclusion of so 
much importance as this. However, as 
it is the only evidence as yet obtained bearing on the question, it 
must be accepted until other data are obtained. It is possible that 
one other date is given by Maudslay in plate xix, part 5, in an 
inscription found at Chichen-Itza and shown in our figure 1G8. Pos- 
sibly this may be intended for ? Ahau 2 Cumhu, and if so -would be 
the second day of the month in Cauac years, and in accordance with 
the Yucatec calendar. It must be admitted, however, that this is 
very doubtful. It will be noticed that in the inscription from Xca- 
lumkin the glyph B3, to the right of the supposed Chicchan sym- 
bol, consists of two faces, hence is presumabl} 7 double, and over each 
are two large dots. If the first or left one be intended for a month 
symbol, there is still correspondence with the Yucatec calendar, as 
Chicchan is the second day of the month in Kan years. However, it 
must be admitted that as yet we are unable to solve the problem. 

In regard to the types of the glyphs their nearest approach is to 
those on Stela P, Copan (see Maudslay, plate lxxxviii, part 4). 

Fjg. 168. Two symbols from a 
Chichen-Itza inscription. 





In figure 169 is given a copy of an in- 
scription on Stela 6* at Copan. As the 
photograph of this inscription has been 
kindly furnished "by E>r Saville, who may 
intend to publish further notice of it, I 
shall notice only the initial series. 

This series is as follows (the great cycle 
being neglected) : 9-12-10-0-0, 9 Ahau 18 
Zotz. The chuen and day symbols are too 
indistinct to be determined by inspection. 
The s}~mbol of the day 9 Ahau is the right- 
hand portion of glyph B2; and that of 18 
Zotz is the right-hand portion of glyph B4. 
Changing the 9 cycles, 12 katuns, and 10 
ahaus to days gives the following result: 


9 cycles 1,296,000 

12 katuns 86, 400 

10 ahaus 3,600 

Total 1,386,000 

Subtract 73 calendar rounds- _ 1, 385, 540 



Counting back 460 days from 9 Ahau 
18 Zotz, year 10 Akbal, we are brought 
to 4 Ahau 8.Cumhu, year 8 Ben, the in- 
itial day of Goodman's 54th great cycle. 
The series, as given above, may therefore 
be accepted as correct, and the lower part 
of glyph A2 as denoting chuens, days, 
or at least chuens. Enough of the left 
half of the lower portion of this glyph re- 
mains to show beyond question the sym- 
bol of full count or naught. 

Dr Saville has also presented me with 
photographs of inscriptions discovered at 
Seibal, Guatemala, but these are short 
and contain no initial series. The only 
peculiarity noticeable is the rjrominence 
at this place of the date 3 Ahau 3 Kayab, 
which stands at the head of some of the 
inscriptions. This shows that the calen- 
dar used here was the same as that in use 
at the other points not in Yucatan, to wit, 
that having Akbal, Lamat, Ben, and 
Ezanab as the dominical days. 



»■*■*;.. S3&JKK 

i 1 

I 4 

« ■ 


Li—-...'- ,*w. . 

A B 

Fig. 169. Inscription on Stela 6, 
Copan. From photograph by 



[ETH. ANN. 22 


Reference is made to the inscription on this stone (figure 170), 
which has been frequently noticed heretofore, merely to show the date 
from which the initial series is counted. The series is as follows, 

omitting the great cycle: 8-14-3-1-12, 1 Eb 5 
(month). The month symbol, though distinct, 
r/y^/^yrr'^) is unusual, in fact unique, unless it includes 

the "kin" glyph immediately below, which is 
very probable; in this case it is most like the 
Yaxkin symbol. Reducing the series to days 
and subtracting 66 calendar rounds, we have 
the following result: 


8 cycles 1, 152, 000 

llkatuns 100,800 

3 ahaus 1 , 080 

1 clraen 20 

12 days 12 

Total - 1,253,912 

Subtract 66 calendar rounds- _ _ 1, 252, 680 



Counting forward 1,232 days from 4 Ahau 8 
Cumhu, the first day of Goodman's 54th great 
cycle, Seler's "normal date," we reach 4 Ix 10 
Xul. This is wrong; but by counting forward 
from 4 Ahau 8 Zotz, the first day of Goodman's 
53rd great cycle, we reach 1 Eb 5 Zac, which 
agrees with the inscription so far as the day 
and day number and the day of the month 

M / j&y-\ M\ I are coneerne( l> but still leaves the doubt as to 

the month. This result also agrees with Good- 
man's tables and his interpretation of this 
series (page 148). Assuming it to be correct, 
we find the terminal date to be 618,088 days 
back of or anterior to the "normal date," 4 
Ahau 8 Cumhu ; and the commencement of the 
10th (Goodman's 9th) cycle of the 54th great 
cycle stands 1,296,000 days after this normal 
date; hence the time of inscribing the series 
on the nephrite stone (assuming the terminal 
date to indicate this time) was 5,244 years an- 
terior to the beginning of the 10th cycle, the 
anterior limit fixed by Seler for the date of 
the inscriptions. However, it must be remembered that this calcula- 
tion is based on the theory that the series on this stone falls in one of 
the three great cycles tabulated by Goodman. This theorj 7 , as is 

Pig. 170. Inscription on the 
nephrite stone in the Ley- 
den Museum. 


apparent from what has been shown in this paper, is not entirely sat- 
isfactory. If the count be backward from 1 Eb 5 Yaxkin, the appar- 
ent date of the inscription, we reach, as the beginning day of the 
series, 4 Ahan 13 Cumhu, which is the initial day of Goodman's 11th 
great cycle; but it must be remembered that 4 Ahau 13 Cumhu will 
appear again and again, in fact hundreds of times, and at much more 
recent dates than this immense stretch of more than 224,500 years. 
Moreover, it is proper to bear in mind the fact that Goodman's list 
of 73 great cycles covers the list of ahaus or 360-day periods com- 
mencing with 4 Ahau; hence any date having 4 Ahau will be found 
somewhere in it. 


Although the following tables are given in my previous paper, it is 
thought best to reinsert them on the following pages (303, 304) for 
the convenience of readers disposed to test the calculations made in 
this paper. 



[ETH. ANN. 22 











h « m ■>* i; c i' OC' c o ^ 

o? or -t- ia © *~ x © © 

t— It— I t— I t-H t— i t-h t— It— I CJ 





£ ■ 








t- x © © t-h oi co th o co -+ io © i~ x © © t-h c? co 

T— I TT— I T— t 1— t tH 1— I T—i T-H 

or T-t <?■} or ^ L'r © i- 'X © o t- i c> or t- i c« or ■«* io © 

«o *.- X' © o t-i <o* or h cm or ^ »o o i- x © 


o> co th c> or ^ io © i- x © © t-h ei co t-h <>? co ^ lo 

T-i T-H H H lH -I— I 

1C «0 I- X 1 flJ.OH'«06H(X,85'<Ji , Oei>Q0®OTH 

T-H TH T-H T-H tH T— I 

t-i o"> or th c* or -+i io © r> x © © t-h oi or h o. or -^ 


■^ ic © J> x os © th <?■) CO T-H CQ 00 ^* JO © i- X © © 

T-H T-H T-H T-H T-H 

© t-h o') or t-h o'j co -t)h ic © t- oo © © t-h o or t-h c"? co 

T-HtHt — IT-H 1 — ll — ItHt-H 

co Tf lc © t' x © © t-h o"j co t-h o? co ^ l-t © i-- 00' © 

T-H T-H T— T-H 

© © th oi co t-h o^ or ^ ic © i- x © © t-h o? or th <o"? 

t-Ht-Ht-Ht-H t-H t-h t— I t-H 

o> co -hh lc © i> x © © th o* co t-i ci or ^ io © *> x 

T-H T-H T-H T-H 


x © © t-h o> or t-h (o^ co ^ i-o © i.~ x © © t-i o> or t-h. 

1 — It— It— it— I . T— I t— i t— I t-H 

t-h o"* co ^f 10 o i- x © © t-h ci or t-h ci or th-i ic . © *.- 

S £ S « J r. O 

N to" -£ g ►- -^ * .3 -B iS jgr- o ^ ^o <g 

•rH r"i "3 

a r" 75 

►> u_| 'H TO 

^ S o o 

- -i 



^ r^ cs esj a ^ 

eg M 

TO . r- 1 

^ O O H O <j 

S r*! 


H rC 

• pH TO r* 

■■-l rH r-l r - 

a . S | 9 




C3 r>| 


« ^ 5 ^ s 

C3 ^ CD rd h2 CD 

Hq g O Q H PQ 


fl ct a r- 

_, C5 fl «3 rH W 

,2 S C8 N c8 fl fl 



^ 3 •§ 



,-, .pH TO CM TO >— ■< H M ' > i rH •!-! u_| 



■a f 

O rn w 
r-^ rH .^ 

r^ es ^ 


TO r-l 



ro -a 

c3 a 



d c? .3 


i^ * rfl -H 5 Ci £ -rCirQ CD ^^j^^ N TOra fl 




Working tables 


r rounds 


• rounds 


18, 980 


398, 580 


778, 180 

61 1,157,780 


37, 960 




797, 160 

62 1.176,760 


56. 940 


436, 540 


816, 140 

63 1.195,740 


75, 920 


455, 520 


835, 120 

64 1,214,720 


94, 900 


474, 500 


854, 100 

65 1,233,700 



493, 480 



66 1,252,680 




512, 460 



67 1,271,660 







68 1,290.640 




550, 420 



69 1,309,620 


189, 800 




949, 000 

7(1 1,328,600 


208, 780 


588, 380 



71 1,347,580 


227, 760 


607, 360 


986, 960 

72 1,366,560 


246, 740 


626, 340 



73 1,385,540 


265, 720 


645, 320 



74 1,404,520 


284, 700 


664, 300 



75 1,423,500 




683, 280 



76 1,442,480 


322, 660 


702, 260 



77 1,461,460 







78 1,480,440 


360, 620 


740, 220 



79 1,499,420 


379, 600 





80 1,518,400 









144, 000 












432, 000 




28, 800 








720, 000 




43, 200 


864, 000 




50, 400 






57, 600 






64, 800 


1 , 296, 000 


3, 600 














86, 400 


1 . 728, 000 




93, 000 







1 1 

2. OKI, 000 


:». loo 


1 OS, 000 


•2. 100,000 








6, 120 


122, 100 






129, COO 








2, 736,000 




I 11.000 



22 ktii <>4 



[Roman numerals in small capitals are numbers of plates.] 


Abnaki Indians, observations on x 

Activities, tribal, definition of x 

Adamana, Ariz., ruined pueblo near 19,136 

Adams, Mr, excavations at Epley's ruin by. 171 
Adams's cave, Graham mountains, visit to. 188 

Adobe, use of, in pueblo buildings 29, 176 

Age, importance of, among savages... xxv-xxvi 

Ahau, in Mayan chronology, value of 204,234 

symbols for, discussion of 222-223, 265 

figures showing 267 

Alaska, field work in xiv 

Algonquian Indians, field work among xii 

Algonquian vocabulary, work on xxxii 

Amphora from cave in the Nantacks 189 

Anderson, Peter, pueblo ruins on farm of.. 173 
Andrade, Eduardo J., acknowledgment to. xii 
Animal effigies, finding of, in pueblo ruins. 109 

Animal figures on Pueblo pottery 71, 72, 

145, 153, 154, LXb 
Animal forms of pottery from pueblo ruins, 

features of 66-69, 133 

figures showing 67, 68, 133 

Animal remains from pueblo ruins, collec- 
tion of 26, 164 

list of species of 110 

Antelope, pueblo implements made from 

bones of, descriptions of 94, 95 

Apache basketry, collection of xxxix 

Apaches, attacks by, a probable cause of 

abandonment of pueblos 20 

Arachnid figures on Pueblo pottery 81-83 

Aravaipa valley, field work in xiii 

Archeological explorations in Arizona 1-195 

Architecture, Pueblo type of, evolution of. 193-195 

Areneiios, field study of xi 

Arizona, archeological explorations in, ac- 
count of 1-195 

archeological explorations in, plan 

of, for 1897 121-123 

collections made in, by Dr Russell xxx 

distribution of ancient pottery in, map 

showing lxx 

field work in ix,xi, xiii 

office work on collections from xvii,xxiii 

petrified forest in, habitations in 135-136 

pottery in, distribution. of 192-193, i.xx 

effigy vases from, description of 189-192 

Two summers' work in Pueblo ruins in, 

paper by J. Walter Fewkcs 1-195 

characterization of paper on xliii 


Armlets from Chevlon, figures showing 89, 90 

Arrow heads of volcanic glass 183 

Arrow points of stone from pueblo ruins . . . 103 
Arrow polishers from Pueblo Viejo ruins, 

description and figures of 182, 183 

Arrows and bow, finding of, at Chevlon 100 

Art, primitive, .symbolic types of, among 

peoples of Walpi, study of. xvii-xviii 
Asbestos, use of, at Chaves pass pueblos.. 110-111 

Asphalt found at Chevlon ruin Ill 

Aspiroz, Manuel de, acknowledgment to. . . xi 

Avian figures. See Bird designs. 

Awls, bone, from pueblo ruins, description 

and figures of 94 

Ax, stone, from Homolobi, figure showing. 102 

Babacomori valley, field w r ork in xiii 

Balls, stone, from Pueblo Viejo ruins 183 

Bark, birch, study of Indian use of xxii 

Basketry, Apache, collection of xxxix 

Pueblo, types of 98-99 

Beads, bone, finding of, at Chaves pass ruins. 95 
Beads, shell, from ruined pueblos on Little 

Colorado river, description of.. 92-93 
Bear design on food bowl from Four-mile 

ruin, description and figure of. 153, 154 
Bear's paw designs on pottery from pueblo 

ruins, figures showing 73, 155 

Bells, copper, from pueblo ruins. . . 50, 111, 162-163 
Belt frame, stone used in, from pueblo ruin, 

description of 159-160 

figure showing 160 

Birch bark, study of Indian use of xxii 

Bird designs on pottery from pueblo ruins, 

descriptions of . 73-81, 115-118, 146-148 

figures showing 74, 

76, 77, 78, 79, 80, 81, 115, 116, 
117, 118, 132, 146, 147, 148, 154 

Bird effigy, clay, from pueblo ruins 109 

Bird eggs, finding of, in grave in pueblo 

ruins in Arizona Ill 

Bird-form vases and vessels from ruined 

pueblOS Of Arizona, features of. 66-69 

figures showing 67,68 

Bird-snake vase from Chevlon, cut show- 
ing 68 

Birds, stone, from Woodruff butte, descrip- 
tion and figures of 1:55 

Bison-like figure on food howl from Chev- 
lon, description and figure of... 72 




[ETH. ANN. 22 

Black falls of Little Colorado river. See 

Little Colorado river. 
Black ware from pueblo ruins, reference 

to 59,179 

Black and red ware from pueblo ruins 60, 


Black and white ware from pueblo ruins, 

reference to 59-60, 180 

plates showing xx,xxxii,xxxiv, 

xxxv, XXXVI, XLI 
Black, red, and white ware from pueblo 

ruins 60-61, xxi-xxv, 


Boas, Franz, office work by xxxii 

Bodkins from pueblo ruins on Little Colo- 
rado river, finding of 94 

Bone implements from pueblo ruins, de- 
scriptions of ... . 93-95, 134, 164, 166-167 

figures showing 94, 95, 166 

Bonita creek, Arizona, cliff houses on 187 

Bow and arrows, finding of, at Chevlon 100 

Bowls from pueblo ruins in Arizona, bear 

design on , cut showing 153 

bear's paw design on, cut showing 155 

bird designs on 76-78, 115-118, 146, 147, 148 

cuts showing "... 74, 76, 77, 78, 79, 

80, 115, 116, 117, 118, 132, 146, 147, 148, 154 

butterfly design on, description of 148 

cut showing 149 

dragon-fly design on 147 

exterior decoration of 152-155 

cuts showing 132, 153, 154, 155 

feather design on, cut showing 151 

features of 61 , 113, 133 

figures and plates showing 114, 

131, 132, 141, 151, xxin, ' xxiv, 


fr,og or lizard design on 132 

gaming-cane design on, cut showing . . . 118 
geometrical designs on 83-84, 85, 152, 155, 


human face on 71, xxvn 

human figures on . . . 143, 144, 145, xxiv, xxva 

quadruped designs on 72, 145, 153, lx& 

rain-cloud designs on, character of. 73-74, 156 

cuts showing 74, 157 

reptilian design on, description of 115 

cut showing 114 

spider and sun emblem on, description 

and figure of 81-82 

spiral design on, views showing . . . 151, lxiv 

sun emblem on, cut showing 150 

Bowlder sites of Pueblo Indians, uses of 177 

Bracelets of shell from pueblo ruins, discov- 
ery of 90-91 

Breath-feather designs, occurrence of, on 

] lottery from Old Shumopovi . 115-116 
Brinton, I). G., on Cakchiquel cal- 
endar 275-277, 279 

British Columbia, field work in ix, xii 

linguistic materials collected in, work 

on xxxii-xxxiii 

Broad house. See Kintiel. 

Brown and red ware from pueblo ruins . . . 58-59, 

Brown, red, and yellow ware from pueblo 

ruins, plate showing xxv 

Buena Vista ruin, description of 171-173 

plan of, plate showing lxvi 

Buffalo. See Bison. 

Building materials used by Pueblo 

Indians 176, 177 

Burro spring, ruins near, reference to 55 

Butterfly design on pueblo pottery, descrip- 
tions of 81, 148 

figure showing 149 

Butterfly or moth, Pueblo symbol for 105 

Caborca, Tepoka Indian habitations near, 

traces of xi 

Cakchiquel calendar, discussion of 275-282 

names of days of, table showing 278 

year of, number of days in 279-282 

Cakwabaiyaki ruin. See Chevlon. 
Calendar round in Mayan chronology, defi- 
nition of 206 

Calendar systems of Mayan Indians, paper 

by Cyrus Thomas on 197-305 

characterization of xliii 

Calendar tables of Mayan chronological 

system 1 ,... 304-305 

Calendric terms in Mayan chronology, defi- 
nition of 204, 234 

California, field work in ix 

stone objects from, purchase of xxxix 

Canada, field work in ix, xiii 

See also British Columbia. 
Canals for irrigation, used by Pueblo 

Indians, remains of 178 

Canes, gaming, from Chevlon, cuts showing 101 
Canoes, birch-bark, Indian use of, study 

of xxii-xxiii 

Canteens from pueblo ruins, forms of 64 

Carapaces of turtles from pueblo ruins, 

description of 95 

Carrizo, Ariz., ruins near, mention of 19 

Carroll, John J., work of xii 

Casa Grande, people of, relations of, to 

people of Verde valley 34 

Castafieda (de Nagera), P. de, on cremation 

among Pueblo Indians 176 

on stone balls found in pueblo ruins . . . 183 

Cavate ruins near Flagstaff, classes of. 35 

descriptions of, published 35 

New caves of 37-38 

Old caves of 36-37 

plan of 37 

Turkey Tanks caves of 38-39 

views of i-vi 

Cave, sacrificial, in the Nan tacks, collection 

of objects from 188-189 

Graham mountains, description of 187-188 
Cemetery at ruin a, Black falls of Little 

Colorado river, description of . . 53-54 
objects obtained from, description of . . . 54 
Cemetery, modern, at Hopi pueblos, Arizona, 

view of lii 

Cemeteries, pueblo, objects found in. See 
the various ruins. 

PART 1] 



Central American tribes, language of, study 

of xxxi-xxxii 

Cephalopoda, fossil, veneration for, among 

Pueblos 107-108 

Ceramic ware. See Pottery. 

Ceremonial chamber at Four-mile ruin, 

mention of 173 

Ceremonial slabs, stone, from Pueblo Viejo 

ruins, cuts showing 185, 186 

Chaves pass ruins, asbestos found at 110-111 

bone implements from, similarity of, 

to those of Four-mile ruin 164 

bone implements from, views of 94, xlv 

bowls from, bird designs on, description 

and figures of 75, 76 

geometrical designs on, description 

and figure of 84 

quadruped design on, description 

and figure of 71, 72 

views of xxviii, xxix 

building material used at 33 

burials at 34 

cloth fragments discovered at 97 

collections at, study of xviii 

copper bell from, cut showing Ill 

crania from, collection of 34, 110 

date of work on 17 

description of 33 

dog's skull found at, features of 27 

dippers from, plate showing xxxiv 

gorget from, cut showing 86 

horn objects from, reference to 96 

human crania from, collection of 34, 110 

kaolin cup from, cut showing 96 

location of 23, 32 

matting from 98 

metates found in graves at 104 

mosaic frog from, view of xliv 

mosaic ornaments found in 85, 86-87 

pottery from, similarity of. to that 
from Verde and Gila-Salado val- 
leys 34 

views of xxviii, xxix, xxxiv, xxxvi 

prayer sticks from, description of 100 

shell ornaments from 88, 89, 90, 91 , 92 

skulls from, collection of 34,110 

specimens collected at, number of 18 

stone implements found at 102-103 

Cherokee Indians, collections among xxxix 

field work among xii 

myths of, work on xxxiv-xxxv 

Chevlon ruins, armlets from, figures of 89,90 

asphalt found at Ill 

ha-ketry from, coiled type, cut show- 
ing 99 

Oraibi type, cut showing '.in 

bird figures on pottery from 7:'. 

hird-shaped vessel from, figure show- 
ing 68 

bird-snake vase from, figure showing.. 68 

hone awls from, reference to 94 

bow and arrows from, reference to 100 

howl from, bison design on, description 

and figure Of 72 

bowls from, bird figures on, description 

and figures of 71. 77-7s. 7'.» 

Chevlon ruins, bowls from, geometrical de- 
signs on, description and figure 

Of 83-84, 85 

rain-cloud symbols on, character 

of 73-74, 156 

figure showing 74 

views of xxviii, xxix, xxxiv, xxxv, 


burial place at 32 

collections from, study of xviii 

crania from hq 

cups from, figures showing 65, 66 

date of work on 17 

dippers from, views of xxxiv 

duck-shaped vessel from, figure of 67 

ear pendants from, cuts showing 86, 87 

human crania from, collection of 110 

ladle from, figure showing 62, 63 

lignite ornament from 87 

location of 23,30-32 

matting from, abundance of 97, 93 

metates found in graves at 104 

mosaic ornaments from 85, 86 

mug from, figures showing 66 

objects obtained from, number and 

character of 18, 23 

ornaments, mosaic, found at s;>, si; 

ornaments, shell, from 88, 89, 90, 91, 92 

prayer stick from, exceptional charac- 
ter of 100 

rattle from, made of Conus shell, cut 

showing 91 

shell objects from, figures showing 92 

skulls from no 

stone implements found at 102, 103 

stone slab, with rain-cloud design, 

from 104-105, xlvi 

turtle carapaces from «).">. 96 

vases from, views of xx, xxxvm, x 

Chichen-Itza, inscription at, symbols from; 

discussion and figures of 300 

Chickahominy Indians, remnants of, dis- 
covery of xii 

Chimney-like structure at ruined pueblo 

in Arizona, description of 49 

plate showing \ v 

Chiricahua mountains, field work. in xiii 

Chronological calendar of Mayan Indians. 
Sec Mayan. 

Chuen, in Mayan chronology, value of 204 

symbol for 223 

Cienega, pueblo ruin near 187 

Cinerary urns from Pueblo Viejo, descrip- 
tions of 173,181-182 

plates showing lxviii, lxixc 

Cipias pueblo, location of sa 

Citadel (the) at Black fallsof Little Colo- 
rado river, building materia] of. II 

view from vn 

view of 13 

Clear Creek canyon, cliff houses in, refer- 
ence t > 20 

Cliff houses in Arizona, character of 1 s7 

discovery of. by Dr Russell xiii 

cloth from pueblo ruins, fragmentary speci- 
mens of , preservation of 97 

Cloud emblems. Set Rain-cloud designs. 



[ETH. ANN. 22 


Cloud people, clans of . . 24 

Cocopa Indians, collections made among, xxxix 

field study of xi-xii 

habitat of xix-xx, xxvii 

girls' puberty feast among xxviii-xxix 

marital customs among xxviii 

mortuary observances among xxi, xxviii 

subsistence of, method of xx-xxi 

Coiled type of basketry from Chevlon, cut 

showing 99 

Coiled vase from Kintiel, cut showing 130 

Collections made during the year, detailed 

account of xxxviii-xxxix 

Color, classification of ware from pueblo 

ruins by 58-61, 179-180 

Consanguinity, importance of, among 

savages xx v 

Conus shells, Pueblo articles made of 91 

Copan, C. A., face numbers from Mayan 

monuments at, discussion of . 221-225 
face numbers from Mayan monuments 

at, plates and figures showing. . 224, 
great cycle symbols on inscription at, 

figures showing 266 

inscriptions on Mayan monuments at, 

discussion of 221-225, 254, 301 

plates and figure showing 224, 225, 301, 


terminal dates of, significance of. . 290-299 
Copper bells from pueblo ruins in Arizona, 

descriptions of 50, 111, 162-163 

figures showing Ill, 162 

Cortesian codex, references to 252, 253 

Crania, human, from pueblo ruins, collec- 
tion of 34,110,134 

Cremation, practice of, by Pueblo Indians 

of Pueblo Viejo 175-176 

Crescent-shaped figures on articles from 
ruined pueblos, forms and vari- 
ations of 89-90 

Cross, Foliated, at Palenque, Central Amer- 
ica, face numerals on, value of. . 219 
Cup from Kintiel, description and figure of. . 133 
Cup, kaolin, from Chaves pass ruin, cut 

showing 96 

Cups from ruined pueblos on Little Colo- 
rado river, cuts showing. . 64, 65, 66, 96 

forms of 65-66 

See also Dippers; Mugs. 
Cushing, F. P., Kintiel ruin discovered by. . 124 
Cycle in Mayan chronology, numerical 

value of 204,234 

symbol for, character of 222 

figures showing 268 

Cycle, great. See Great cycle. 

Cyclopedia of Native Tribes, progress of 

work on xxxvii 

Cylinders found in pueblo ruins, problem- 
atic use of 96-97 

Dancer (masked), face of, on pottery from 

Shumopovi Li,115 

figure of, on bowl from Four-mile ruin, 

cut showing 144 

description of 143 

Deer, implements made from bones of, from 

pueblo ruins, descriptions of... 94,95 
Devol, W. S., account by, of cliff houses on 

Bonita creek, reference to 187 

Diccionario de Motul, work on translation 

of xxxiii 

Dippers from pueblo ruins 130, 133 

handle of, cut showing 131 

views of 65, 130, 133, xxxiv 

See also Cups; Mugs. 

Disk of turtle shell from Chevlon 96 

Disks of clay from Kintiel 133 

Disks of various materials from pueblo 

ruins, list of specimens of 106 

problematic use of 96-97 

Dog, skull of, found at Chaves pass ruin, 

features of 27 

Doney, Benjamin, objects collected by, 

from Arizona pueblos 39, 49-50 

Dorsey, G. A., pottery found at Homolobi 

by 28 (note) 

Dragon-fly design on pueblo pottery, cut 

showing 147 

reference to 81 

Dresden codex, figures from, illustrating 

Mayan chronologic system 217, 

' 259, 282, 297 

plates from, copies of lxxIx-lxxxii 

discussion of 237-239, 243-248, 286-290 

vigesimal system used in 234 

Duck-shaped vessel from Chevlon, cut show- 
ing 67 

Duff, N. Francis, on prehistoric ruins of Rio 

Tularosa 180 

Eagle, use and ceremonial burial of 28 (note) 

Ear pendants from pueblo ruins, cuts show- 
ing 86,87 

discovery of 86, 91 

Effigies, animal, finding of, in pueblo ruins. 109 
Effigy vases from southern Arizona, descrip- 
tion and figures of 189-192 

Eggs of birds, finding of, in grave in pueblo 

ruins on Little Colorado river. . Ill 

Eighteen-mile spring, Arizona, ruin at 127 

Elsmereland, visit of Robert Stein to xiv 

Emmons, G. T. , collection of obsidian blades 

obtained through xxxix 

Emory, Lieut. W. H., on building materials 

used at Pueblo Viejo 177 

on Pueblo Viejo valley in 1846 174 

Epley's ruin, Pueblo Viejo valley, descrip- 
tion of 171 

objects obtained from 171 

pottery from, views of lv, lxviii 

Esthetology , office work in x v-xix 

Ethnic system of Bureau, features of ix-x 

Ethnobotanyof the Hopi Indians, studiesin. 17 
Ethnology, descriptive, office work in . . xxxvii- 

Everett, Willis E., field observations by, on 

Alaskan tribes xiv 

Face numerals in Mayan calendar systems, 
discussion of and figures show- 
ing 204-226, 263-268 

PART 1] 



Feather decoration on pottery from pueblo 

ruins, description of 149 

figure showing 151 

See also Bird designs. 
Feather symbolism on Pueblo pottery, 

occurrence and character of. 73,75-81 
See also Breath feather. 
Fetish, stone, from Pueblo Yiejo ruins, cut 

showing 186 

Fetishes from pueblo ruins, character and 

uses of 107-109 

Fewkes, Walter, office work by .. . xvi-xix,xxiii 
Two summers' work in pueblo ruins in 

Arizona, report on 1-195 

characterization of report on xliii 

work of, on Pueblo mythology xxxvi 

Field work, detailed account of x-xiv 

scope of ix 

Files or saws of stone from pueblo ruins, 

description of 159 

Filfot, wide range of, among American 

aborigines xv-xvi 

Financial statement xlii-xliii 

Finger rings, shell, from pueblo ruins in 

Arizona, discovery of 91 

Flagstaff, Ariz., cavate ruins near 35 

cavate ruins near, descriptions pub- 
lished of 35 

New caves of 37-38 

Old caves of 36-37 

Turkey Tank caves of 38-39 

views of i-vi 

Fleas, sand, plague of, a possible cause of 

abandonment of pueblos 22 

Fletcher, Alice C, field work by, among 

the Pawnees xiv 

office work by, on Pawnee ritual of the 

Hako xxxvii 

record of Hako ceremony obtained by, 

characterization of report on . . . xliv 
Floods, a possible cause of abandonment of 

pueblos on Little Colorado river 22 
Floor and roof construction at Homolobi, 

manner of 110 

Flute ceremony of Pueblo Indians, refer- 
ence to 18 

Foliated Cross at Palenque, face numerals 

on, value of 219 

Food, remnants of, in mortuary bowls at 
ruined pueblos on Little Colo- 
rado river 101 

Food howls. See Bowls. 

Footprints, human, representations of, on 

Pueblo pottery 70, 71 

Forbes, R. H., field work by xii 

For< -i. petrified. See Petrified forest. 
Form, classification of ware from pueblo 

ruins by 61-69 

Four-mile ruin, Arizona, animal remains 

from, collection of p;i 

bell from, description and figureof.. 162 L63 
bone implements from, similarity of, 

io those from Chaves pass i<;i 

bowl from, double spiral design on i;.i 

Set also Four-mile ruin, pottery from. 

Four-mile ruin Arizona, cemeteries of, 

skeletons and mortuary objects 

from 139-140 

copper bell from 162-163 

excavations at, view showing lvii 

food bowls from, decoration of 143-158 

decoration on, figures showing 1 44-155, 157 

views of xxiv, xxv, 


See also Four-mile ruin, pottery 
from, beloxo. 
former inhabitants of, possible reference 

to, by Spanish writers 123 

gourd rattle from, description of 163 

kiva in, mention of 173 

location and general plan of 136-137 

loom stones from, view of i.xv 

mounds at, surface of, view showing... liv 

ornaments from, scarcity of 163 

pitcher from, views of i.ix 

potter's outfit from 158 

pottery from, bear design on 153 

bear's-paw design on, cut showing. 155 

bird designs on 146-148 

cuts showing . . 146, 147, 148, 149, 154, 155 

butterfly designs on 148, 149 

cloud emblem on, figure showing.. 157 

decoration of, character of 142-143 

dragon-fly designs on, cut show- 
ing 117 

feather decorations on 149, 151 

forms of 142 

geometrical designs on, character 

of 150-152 

figures and plates showing . . 151, 152, 


human figures on 143-145 

views of 141, xxiv, xxva 

mortuary, character of 158 

quadruped figures on 145 

plate showing lx& 

rain-cloud designs on, description 

of 155-158 

cuts showing 150, 157 

similarity between that from Pue- 
blo Viejo and _ 1 Jl 

spiral designs on 151, lxiv 

sun emblem on, cut showing 150 

types of, description of 140-1 11 

pottery of Gila type in, occurrence of .. i n 

pueblo ruins near Kil 

room in, description of 137-139 

I 'Ian of lvii I 

upright posts in, cut showing 138 

saucer from, decoration on 142 

stone from, used in bell frame, descrip- 

t ion and figure of 159 160 

stone implements from 159 L60 

stone slabs from, character of 160 162 

views of 161, w 

suburban ovens of L39 

vases from, views of lix, r,xn 

\ esse! from, Sgure of raccoon on, sui 

gestion of 1 r>. 1 Kb 

View of 136 



[ETH. ANN. 22 

Fretwork designs, Pueblo articles bear- 
ing 85, 89-90 

Frog design on bowl from Kintiel, de- 
scription and figure of 132 

Frog, figures of, from pueblo ruins, views 

of 92, x'liv 

Full count in Mayan face numerals, sym- 
bols for, discussion of 210-211, 

211-212, 214, 215 
symbols for, figures showing 216, 217 

Galena, specimen of, from Chaves pass Ill 

Galiuro range, field work in xiii 

Gaming canes, design of, on food bowl from 

Old Shumopovi, cut showing. . . 118 

from Chevlon, cut showing 101 

Gardens, terraced, of Pueblo Indians, 

features of 177-178 

Gatschet, Albert, office work by xxxii 

Geometrical designs on pottery from pueblo 

ruins, character of 83-85, 150-152 

figures showing 83, 84, 85, 155 

preponderance of 69 

Georgia, stone implements from, Steiner 

collection of, purchase of xxxix 

Gesture language, reference to xxx 

Gila valley, field work in . xiii 

pottery from, characteristic color 

of 193 

characteristic type of 1 41 

occurrence of, in Four-mile 

ruin 141 


visit to 121 

See also Pueblo Viejo. 
Gila-Salado basin, pottery from, similarity 
of, to that from Verde valley 

and Chaves pass 34 

Gill, De Lancey, field work by xii 

office work by xxxviii 

Glazed pbttery , ruins furnishing 57 

Gnats, plague of, a possible cause of aban- 
donment of pueblos 22 

Goodman, J. T., interpretations of Mayan 

calendar by, reference to xliv 

on Mayan inscriptions and chronol- 
ogy 203-305 (often) 

summary of work of, on Mayan inscrip- 
tions 261 

Gorgets from pueblo ruins in Arizona, dis- 
covery of 86, 91 

figures shoAving 86, 92 

Gourd rattle from pueblo ruins 163 

Graham, Mount, pueblo ruins near 187 

Graham mountains, sacrificial caves in, 

visit to 187-188 

Grand era, numerical value of, in Mayan 

chronology 234 

Grave slabs at Chevlon and Homolobi, per- 
forations in 106 

Gray ware, Pueblo Viejo ruins, character 

(»!' 180 

Great cycle in Mayan chronology, discus- 
sion of 262-275 

numerical value of 204,234 

symbols for, figures showing 264, 266, 298 

Great cycles of Mayan chronology, initial 

days of, discussion of 236-248 

initial days of, tables showing . . . 236, 255, 256 

Hako ritual of the Pawnees, work of Miss 

Fletcher on xiv, xxxvii, xli v 

Hales, H., pottery collected by 180 

Haliotus shells, Pueblo articles made of . . . 92 

Hardy, Ariz., ruin examined near 23, 31 

Hatchets, stone, from Pueblo Viejo ruins, 

collection of 183 

Havasupais, cavate dwellings of ancestors 

of . 35-36 

Helmets, use of, in ceremonial dances of the 

Hopis 143 

Hewitt, J. N. B., collections made by, 

among Iroquoian Indians xxxix 

work of, on Iroquoian languages, 

myths, and customs xiii, 

xxix-xxx, xxxii, xxxv-xxxvi 
H-figure on Pueblo pottery, reference to... 181 
Hilder, Frank Frederick, biographic sketch 

of xl-xlii 

death of xxxviii 

work of ". xxxvii-xxxviii 

Hodge, F. W., aid given Dr Fewkes by 120 

office work by ,xxxvii, xxxviii 

Holbrook, Ariz., petrified forests near.. 19 (note) 

pueblo ruins near, visit to 19, 134-136 

Homolobi group of pueblo ruins, animal re- 
mains found at, list of species of 110 

bone implement from £4 

cloth fragments discovered at 97 

collections from, study of xviii 

date of work at 17 

dippers from, figures of 65, xxxiv 

excavations at, by Dr Fewkes xviii 

food bowls from, views of xxiu, 

XXVII, xxix, xxxv, XLIII 
See also, Homolobi, etc., pottery from, 

former inhabitants of 24-25 

jar from , figure showing 67 

location of 23-24 

matting found at 98 

metates found in graves at 103, 104 

mug from, figure showing 65 

pipe clay, fragments of, found at Ill 

pottery from, bird figures and symbQls 

on 73, 75, 76-77, 81 

human face and figures on 71 

spider and sun emblem on 81-82 

views of 65, 67, 


prayer sticks from, form of 100 

roof and floor construction at 110 

ruin 1, bones found at 26 

cemetery at, excavations in 28 

features of 25, 26 

location of 25 

plan of 26 

ruin 2, excavations at 29 

ruin 3, adobe blocks used at 29 

bone implements found at 30 

PART 1] 



Homolobi group of pueblo ruins, ruin 4, 

features of 30 

pictographs at 30 

shell ornaments from 88, 89, 92 

specimens collected at, number of 18 

stone implements from, cuts showing.. 102 

traditional location of 22 

vase from, bird figures on, character of. 75-76 
vases from, views of XXI, 


Honani, ruins near house of, reference to.. 55 
Hopi Indians, abandonment of pueblos on 
Little Colorado river by, causes 

of 22 

esthetic standards among, permanence 

of xxiii 

ethnobotany of, studies made in 17 

evidence from ritual of, as to former 

inhabitants of Homolobi 24-25 

helmets or masks used in ceremonies of 143 
legends of, concerning home of katci- 

nas 114-145 

pantheon of, office work on xvi-xvii 

pottery of, superiority of 129 

Hopi pueblos, exploration of ruins at xviii 

modern cemetery at, view of lii 

ruins between Winslow and, descrip- 
tions of 34-56 

Horn objects from Chaves pass ruins, refer- 
ence to 96 

Hough, Walter, aid given Dr Fewkes by.. 17,120 
reference to explorations by, near Hol- 

brook 135 

Huachuca mountains, field work in xiii 

Hubbell's store, Arizona, ruined pueblo 

near, mention of 127 

Human crania from pueblo ruins in Arizona, 

collection of 34,110,134 

Human effigy vases from southern Ari- 
zona 189-192 

Human figures on pottery from pueblo ruins 

in Arizona, descriptions of 70-71, 

cuts showing . . 63, 70. 1 11, XXIV, xxva, xxvub 
Human head on handle of ladle from Pue- 
blo Viejo ruins, mention of 181 

Implements, bone. See Bone implements. 
Implements, stone. See Stone implements. 
Incineration, practice of, by Pueblo Indians 

of Pueble Viejo 175-176 

Indian portraits by J. H. Sharp, purchase 

of x x x i x 

Initial days of great cycles of Mayan chro- 
nology, discussion of 236-248 

tables showing 236,255,256 

Insect figures on Pueblo pottery 81-83 

Iroquoian Indians, collections made by Mr 

Bewitt among xxxix 

field work among xiii 

languages of, work on xxxii 

mythsof, work on xxxv-xxxvi 

Irrigation, prehistoric, in Pueblo Viejo 

valley its L79 

Isleta, X. M.. visit to 18 

Jack, Mr, aid given In- Fewkes by 39,52 

Jars from pueblo ruins in Arizona, features 

of 61,181 

figures showing 67, 181 

Jenks, Albert E., office work by xxi-xxii 

Johnston. A. K., on building materials used 

in Pueblo Viejo 177 

Joseph, Nathan, collection of obsidian 

blades obtained through xxxix 

Kaolin cup from Chaves pass, cut showing. 96 
Katcinas of the Hopi Indians, legendary 

home of 144 

symbols of, on masks 143 

work on xvi-xvii 

See also Masked dancer. 

Kathlamet texts, work on xxxii 

Katun, a Mayan term, definition of 273-274 

numerical value of 204, 234 

symbols for, features of 222, 223, 265 

figures showing 268 

Keam, T. V., mortuary pottery collected 

atShumopoviby. 69 (note), 113 (note) 

Kinna Zinde, ruin of, description of 134 

Kintiel, age or date of occupation of 126, 127 

inhabitants of, speculations concern- 
ing . . 124, 125 

population of 127 

two pueblos so named 127 

Kintiel ruin, architectural features of 125-126 

bone objects from, mention of 134 

bowl from, bird design on 132, 133 

frog design on 132 

bowds from, cuts showing 131 

cemeteries at 126 

cemeteries east of, skeletons in 128 

clay disks from, mention of 133 

coiled vase from, cut showing 130 

cUp from, cut showing 133 

dipper from, description and figure 

of 133 

with decorated handle, cuts show- 
ing 130, 131 

documentary history of 124-125 

explorations at, purpose of 122 

form and general features of 124-128, LIU 

human skulls from, mention of 134 

investigations at, object of 122 

location of 127 

miscellaneous objects from 133-131 

ovens at, description of 128 

plan of i. in 

pottery from, decoration of 131-133 

general features and form of 129-131 

shrine near, objects collected from... 127-12S 

skeleton from 128 

Stone objects from, mention of 133 

stone objects from shrine near L27-128 

wood objects from, mention of 133 

Kisakobi, (day imitation of PectunculUS 

-hell from 91 

Kiva in Four-mile ruin, mention of 17:; 

Kivas in modern pueblos, mention of 17:; 

Knives, -tone, from pueblo ruins on Kittle 

Colorado river 102-103 

Kokyan wruqti, the spider maid, legends 

concerning 81,82 



[ETH. ANN. 22 

Ladles from pueblo ruins in Arizona, fea- 
tures of 63-64 

figures showing 62, 63 

Landa, Diego de, on Mayan chronology and 

arithmetical methods. 252, 253, 274, 282 

Langley, S. P., letter of transmittal to iii 

Language, development of xxx-xxxi 

Leal, Fernando, acknowledgment to xi 

Le Plongeon, A., figure cited from " Queen 

Moo " of, reference to 253, 299 

Leyden stone. See Nephrite stone of Ley- 
den museum. 
Lignite ornaments from pueblo ruins, char- 
acter of 87-88 

figures showing 87, 88 

"Lines of life" on Pueblo pottery, cut 

showing 83 

reference to 90, 181 

Linguistics, office work in xxx-xxxiii 

Little Colorado river, Arizona, pueblos on, 

causes of abandonment of . . . 20, 21, 22 
Little Colorado river ruins, animal re- 
mains from 110 

basketry from 98-99 

Black falls ruins , 39-55 

group a, description of 42-47 

ruin a, description of 43-44 

ruins b, c, d, description of 44 

ruins E, F, description of 44 

ruins G, H, I, J, description of .. 44-45 

ruin g, view of viii 

ruins H and j, view of ix 

ruin j, plan of 45 

views of x, xi 

ruin K, description of 45-46 

ruin L, description of 46 

ruin M, description of 46-47 

the Citadel, description and fig- 
ure of 43 

view from vn 

grbup B, ruin a, description of 47-50 

ruin a, plan of section a of 47 

section of section a of 46 

views of xn-xvi 

ruin b, description of 50 

ruin c, description of 50-51 

group c-, ruin a, description of 51-54 

ruin a, plan of 52 

views of 51, xvii-xix 

ruin B, description of 54-55 

location of 39 

previous exploration of 39 

views of vii-xix 

bone implements from 93-95 

bow and arrows from 100 

building materials of 40, 41 

Burro spring ruins 55-56 

Cavate ruins near Flagstaff 35-39 

Chaves pass ruins 32-34 

Chevlon ruin 30-32 

cloth from 97 

crania from , 110 

disks from 106 

fetishes from 107-109 

food from 101 

gaming canes from 100-101 

general features of 20-22 

Homolobi group 23-30 

Little Colorado river ruins, horn objects 

from 96 

human crania from 110 

matting from 97-98 

objects from 56-111 

ornaments from 85-93 

pigments from 96-97 

pottery from, classification of, by color 

and surface finish 58-61 

classification of, by form 61-69 

color of 192 

decorative designs on „ 69-85 

general features of 56-58 

map. showing 21 

prayer sticks from 99-100 

racial and clan kinship of inhabitants 

of 56 

seeds from 101 

stone implements from 102-104 

stone slabs from 104-106 

turtle carapaces from 95 

Lizard design on bowl from Kintiel, descrip- 
tion and figure of 132 

Logistic language, development of xxxi 

Logs used in construction of buildings at 

Pueblo Viejo ruins, mention of. 177 
Loom stones from Four-mile ruin, yiew of. . lxv 
Lucas, F. A., animal remains from pueblo 

ruins identified by 26,110 

on features of skull of dog found at 

Chaves pass, Arizona 27 

Lund, J. J. , work of xl 

McGee, W J, administrative report by... ix-xliv 
ceremonies among Yaki Indians wit- 
nessed by xxxiv 

ethnologic collection made by xxxviii- 


field work by xi-xii, xxvii-xxviii 

letter of transmittal by iii 

office work by xv-xvi, xxiv 

on Mayan arithmetical methods 282 

Mahler. See Maler. 

Maine, field work in ix, x 

Maize found in pueblo ruins on Little Colo- 
rado river, character of 101 

Maler, Teobert, figure of inscription at Xca- 
lumkin, Yucatan, reproduced 

from photograph by 253 

Mammalian figures on pottery from pueblo 
ruins, description and figures, 

of 71-72, 153-154 

Masked dancer, face of, on pottery from 

Shumopovi 115, LI 

figure of, on bowl from Four-mile ruin, 

cut showing 144 

description of 143 

Masks, use of, in ceremonies of Hopi In- 
dians 143 

Mata, Aurelio, aid given Mr McGee by xii 

Mattaponi Indians, discovery of remnants 

of tribe of xii 

Matting from pueblo ruins, character and 

use of 97 

figure showing type specimen of 97 

Maudslay, A. P., inscriptions from Mayan 

monuments from works of 211, 

217, 218, 221, 224, 225, 227, 229, 230, 
231,254,266, LXXI-LXXV,LXXVIII 

PART 1] 



Maudslay, A. P., work of, In Central 
American archeology, reference 

to xliv, 203 

Maya dictionary, work in transcribing ... xxxiii 
Mayan calendar systems, paper on, by Cyrus 

Thomas 197-305 

characterization of xliii-xliv 

names of days of, table showing 278 

Mayan chronology, ahau symbols in, dis- 
cussion of 265 

figures showing 267 

calendar tables of '. 304-305 

cycle symbol in, discussion of and figures 

showing 234-237, 268 

full count in, symbols for 264 

great cycles in , discussion of 262-275 

initial dates of, table showing 255, 256 

symbols for 264, 266 

katun symbols in 265, 267 

number tables of 304-305 

symbols for numerals in, discussion and 

figures of 210-217, 264 

time periods of, terms designating 204 

value of calendar terms in 204, 234 

Mayan inscriptions, initial series of 203-226 

secondary series of 226-233 

significance of terminal dates in 290-299 

Mayan method of calculation, discussion 

Of 282-290 

Mayan numeral system, figures from Dres- 
den codex illustrating 259 

principles of 259-261 

significance of 290-299 

Menchero, map by, reference to 19 

Metate from pueblo ruin, cut showing 184 

Metates, burial of, with women and girls. 103-104 

mention of 160 

Mexican tribes, language of, study of xxxi- 


Mexico, field work in ix, xi-xii 

hieroglyphic records of, work on xxxvi 

Mica plates, finding of, in pueblo ruins 111 

Migrations of tribes, study of xviii-xix 

Mindeleff, Cosmos, Homolobi group of pueb- 
los identified by 22 

w< >rk of, at Kintiel 125 

Mindeleff, Victor, Kintiel ruin described by 125 

Mishongnovi, Flute ceremony at 18 

Moberly, Jim. work of xii 

Moccasin-shaped jar from Pueblo Viejo 181 

Mogollon mountains, cliff houses in, refer- 
ence to 20 

Molina, Audomaro, work of, on translation 

of Diccionario de Motul xxxiii 

Monolithic animal at Quirigua, figure show- 
ing pari of 211 

great cycle symbol from 266 

Mooney. James, collections made by, among 

Cherokee Indians xxxix 

field work by xii 

work of, on Cherokee myths ... xx.xiv-xxxv 
on Cyclopedia of Native Tribes. . xxxvii 
Morfi, j'adre, manuscript history of Texas 

by, work (.n translation of .. xxxviii 
Mortar from Pueblo Viejo ruins, description 

and figure of 184,185 

Mortars for grinding paint, from pueblo 
ruins on Little Colorado river, 

finding of 104 

Mortuary pottery. See Pottery. 

Mosaic frog from Chaves pass ruin, view 

of XLIV 

Mosaic ornaments from pueblo ruins, char- 
acter of J-5-87 

Moth designs, occurrence of, on Pueblo 

pottery 81 

Moth or butterfly, Pueblo symbol for 105 

Motion pictures of Pueblo industries, 
etc., work of O. P. Phillips 

on xiii-xiv 

Motul, Diccionario de, work on translation 

of xxxiii 

Mount Graham. See Graham, Mount. 
Mug. See Cup; Dipper. 

Mugs from pueblo ruins, features of 65-66 

figures of 64, 65, 66 

Mushroom-shaped implements of stone 

found in pueblo ruins, use of . . . 104 

Myths, Cherokee, work on xxxiv-xxxv 

comparative study of xxix-xxx 

Iroquoian, work on xxxv-xxxvi 

Pueblo, work on xxxvi 

Zuni, work on xxxvi 

Nansemond Indians, remnants of tribe of, 

discovery of xii 

Nantacks, arrowheads of volcanic glass 

from 183 

sacrificial cave in, amphora from, cut 

showing 189 

bowl from, cut showing 188 

collections from 188-189 

human effigy vase from 189 

Natick Dictionary, progress of work on . . . xxxii 
Native Tribes, Cyclopedia of, progress of 

work on xxxvii 

Necrology: Frank Frederick Hiider xl-xlii 

Needle-like implements of bone from 

pueblo ruins, finding of 94 

Nephrite stone of Leyden museum, dis- 
cussion and figure of 302-303 

significance of date on, discussion of. . . 294 
New caves near Flagstaff, Ariz., description 

of 37-38 

entrance to, view of in 

view of n 

New-fire ceremony, use of fetishes in 108 

New Mexico, field w< rk in ix,xiii 

office work on collections from xvii 

New York, field work in ix, xiii 

Nordenskiold, G., bone implements found 
by, in cliff houses al Mesa Verde, 

reference to 95 

features of Mesa Verde ruins described 

and figured by 105, 121 

Xorri-, Bugh, work of, as Papago inter- 
preter xii 

North Carolina, field work in ix,xii 

Nought, in Mayan face numeral-, symbol 

for, discussion of.. Jin -•_>! 1 , 21 1 212,214 
Number tables of Mayan chronological sj - 

tern 304 



[ETH. ANN. 22 

Objects collected. See Specimens. 
Obsidian blades, collection of, obtained 

through Mr Nathan Joseph . . . xxxix 

Office work, detailed report on xv-xxxviii 

general scope of ix, 

Oklahoma, field work in xiv 

Old caves near Flagstaff, Ariz., description 

of 36-37 

view of ' i 

Old Shumopovi ruins, bowl with snout 

from, cut showing 114 

excavations at, stoppage of ._. 112 

general features of 111-113 

pottery from, bird designs on 115-118 

decoration of 114-119 

gaming cane design on 118 

general features and form of 113 

reptilian design on, description and 

figure of 114 

specimens collected at, number of 18 

views of xlviii, xlix, l, li 

Olney, D., pueblo ruins near ranch of 173 

Onondaga Indians, legends of, sociologic 

value of xxix-xxx 

Ontario, field work in ix" 

Oraibi type of basketry, from Chevlon, cuts 

showing 98, 99 

Oral language, development of xxxi 

Ornaments from pueblo ruins in Arizona, 

character of 85-93, 163, 187 

Ovens, Pueblo, description of 128, 139 

Paint mortars from pueblo ruins 104, 184, 185 

Paints, specimens of, from pueblo ruins. . . 96, 184 

Pahos. See Prayer sticks. 

Palenque, inscriptions at, significance of 

terminal dates on 290-299 

Palmer, Dr E., pottery collected by, refer- 
ence to •. 34 

Pamunkey Indians, field work among xii 

Papago Indians, field study of xi 

quatern symbols among xv 

Patki clans, abandonment of pueblos on 

Little Colorado river by, causes 

of 22 

chiefs of 24 

former habitat of . 23, 24 

Patufi people, former inhabitants of Homo- 

lobi 24 

Pawnee Indians, field work among xiv, 

xxxvii, xliv 
Pectunculus shell, clay imitation of, from 

Kisakobi 91 

mosaic work on, views of xliv 

ornaments made of 88-89, 91 

Period symbols in Mayan chronology 222-223 

Petrified forest near Holbrook, ancient 

habitations in 135-136 

Petroglyphs at ruined pueblo at Black falls 

of Little Colorado river 49 

Phillips, O. P., work by, on motion pictures 

of Pueblo industries, etc xiii-xiv 

Philology, office work in xxx-xxxiii 

Photo-portraits of Indians, work on xxxviii 

Piba people, former inhabitants of Homo- 

lobi *. 24 

Pictography, contributions to, by studies of 

pottery from Arizona pueblos . . 69-70 

Pigments, burial of, by ancient Indians, 

sacrificial nature of 96 

pieces of, found in pueblo ruins 96, 184 

Pima basketry, collection of xxxix 

Pinedale ruins, age of 167 

bone implements from 166-167 

buildings of 164-166 

collections from 166-167 

examination of 121 

form of 165 

location of 165 

pottery from, character and decoration 

of 166 

reference to 121 

skeletons from 165 

work at 164 

Pipe clay, fragments of, found at Homolobi . Ill 

Pitcher from Four-mile ruin, views of lix 

Place, Lem., pueblo ruins near house of... 173 
Plumed Snake on pottery from Old Shumo- 
povi, description and figure of. 114, 115 
Polishing stones from ruined pueblos on 

Little Colorado, varieties of 103 

Potter's outfit from Four-mile ruin, finding 

of 158 

Pottery, ancient, in Arizona,- distribution 

of, map showing > lxx 

influence of environment on 122 

Pottery from pueblo ruins in Arizona, classi- 
fication of, by color and surface 

finish 58-61, 179-180 

classification of, by form 61-69 

coarse, decorated, figure showing 65 

reference to 58 

colors of, black, reference to 59 

black and white, features of . . . 59-60, 180 

gray, features of 1 80 

red, reference to 180, 192 

red and black, reference to 60 

red and brown, reference to 58-59 

red, black, and white, reference 

to 60-61 

white and green, reference to 61 

yellow, reference to 59, 192 

decoration of 69-85, 

114-119, 131-133, 142-143, 166, 180-182 

arachnid designs on 81-83 

animal figures on. . 71, 72, 145, 153, 154, lx6 

bird designs on 73-81 , 11,5-118, 146-148 

figures showing 74, 75, 

76, 77, 78, 79, 80, 81, 115," 116 
117, 118, 132, 146, 147, 148, 154 

butterfly designs on 81 , 148 

figure showing 149 

feather designs on 73, 75-81, 149 

figures showing 151 

geometrical designs on. 83-85, 150-152, 155 

human figures on, character of 70-7 1 , 


cuts showi tig 63, 

70, xxiv, xxv«, xxvnb 

insect figures on, features of 81-83 

quadruped figures on, descriptions 

and figures of. . 71, 72, 145, 153, 154, LX& 

form and general features of 56-58, 

113, 129-131, 142, 179-180 

PART 1] 



Pottery from pueblo ruins in Arizona, Gila 

type of 141, 193 

glazed, ruins furnishing 57 

mortuary types of, character and treat- 
ment Of 158 

surface finish of, glazed 57 

polished, decorated, features of. 58-61, 180 
polished, undecorated, features of. 58, 180 

rough, decorated, character of 58,179 

figure showing 65 

undecorated, rough 58, 1-10, 179 

Powell. J. W. , field work by x 

office work by xxiv, xxx-xxxi, xxxiii 

on inhabitants of cavate dwellings near 

Flagstaff 35 

Prayer sticks, absence of, from Four-mile 

ruin 163 

remnants of, from pueblo ruins in Ari- 
zona, description of 100 

Prehistoric migrations of tribes, study 

of xviii-xix 

Property, report on xl 

Publications, Avork on xxxviii 

Pueblo, definition of the term 193 

Pueblo architecture, types of, evolution 

of 193-195 

Pueblo Grande. See Kintiel. 

Pueblo Indians, building materials used 

by 29,176,177 

motion pictures of ceremonies of xiv 

myths of, w r ork on xxxvi 

origin of xix 

Pueblo pottery, distinctive character of , in 

different regions 122-123,142-143 

See also Pottery. 
Pueblo ruins in southern Arizona, explora- 
tions in, paper by J. Walter 

Fewkes on 1-195 

paper on, by Dr Fewkes, characteriza- 

t ion of xliii 

Pueblo Viejo valley, ancient people of, cre- 
mation of (lend by 175-176 

Buena Vista ruins in, description of. . 171-173 

changes in, since 1847 174-175 

Epley's ruin in, description of 171 

former population of 175 

history of 168-169 

irrigation (prehistoric) in, practice 

of ]7n-17«> 

1< >cation of His 

pueblo ruins in, architecture of 176-177 

arrow polishers from 1*2,183 

howls from, views of i.xvii, i.xix 

building materials of 177 

distribution of 170 

effigy vase from L91-192 

general feature- of 168-170 

jar< from, moccasin or slipper- 
shaped 181 

ladle from, human head on is] 

metate from, cut showing 184 

mortars from, description of L84 185 

pigments from, finding of L84 

[lottery from, black and white ware, 

decoration of 180 182 

form, color, and finish of 17'.> 180 

Pueblo Viejo valley, pueblo ruins in, pot- 
tery from, rough ware, undec- 
orated. character of 179 

similarity between that from 

Four-mile ruin and ill 

reference to 173 

shell ornaments from, character of. 187 
stone objects from, description of. 183-186 
vases from, description of. 181-182, 191-192 

views of 191.LXVH, i.xvin,LXix 

visit to 121 

terraced gardens (prehistoric) in 177-178 

Pueblos of refuge, mention of 165 

Pueblos of the Southwest, dual character 

of many of the 165 

Quadruped figures on pottery from pueblo 

ruins, descriptions of 72, 145, 154 

cuts showing 71, 72, 153, LXb 

Quartz crystals, probable use of, by Pueblo 

Indians 186 

Quatern symbols, range and diversity of. xv-xvi 
Quirigua, face numerals of initial series of 

inscriptions at 203-219 

face numerals of inscriptions at, second- 
ary series of 226-233 

Stela A, discussion of 215-217, 232-233 

plate showing i.x.xv 

Stela C, discussion of 217-219, 233, 257 

figure showing 217 

Stela E, discussion of 228-232 

figures and plates showing 229, 

230, 231, lxxii, LXXVIII 

Stela F, discussion of 203-212, 226-228 

figure and plates showing 217, 

lXxi, lxxiii 

Stela J, discussion of 212-215 

plate showing lxxiv 

inscriptions at, great cycle symbols on, 

figures showing 266 

figures and plates showing 211, 

217, 218, 221, 224, 225, 227, 229, 


terminal dates of, significance of.. 290-293 

Rabbit people, former inhabitant- of liomo- 

lobi 24 

Raccoon, suggestion of figure of, on vessel 

from Four-mile ruin 145, LX& 

Rain-cloud people. See Patki. 

Rain-Cloud symbols on objects from pueblo 

ruins in Arizona, description of. . 7:i- 


views of 71,80,81,157, XLVI 

battle, L r ourd, from pueblo ruin-, descrip- 
tion of 163 

shell, from pueblo ruin-, figure show- 
ing in 

bed ware from pueblo ruins, character of. . 1-n 
Red and black ware from pueblo ruin- 60, 

XX XIV, XXX VI, X X \ I \ 

Red and brown ware from pueblo ruin-, 

reference to .- ■ 9 

views showing xwi \\i\ 

Red and white ware from pueblo miii-. 

view - showing \\\v 



[ETH. ANN. 22 

Red, black, and white ware from pueblo 

ruin-, reference to 60-61 

views showing xxi-xxv, xxxiii, xxxiv, 


Red, brown, and yellow ware from pueblo 

ruins, views showing xxv 

Reed. Mr, aid given Dr Fewkes by 39 

Reptilian form on bowl from Shumopovi, 

description and figure of 114, 115 

Reservoir at ruined pueblo on Little Colo- 
rado river, description of 49 

view of xvi 

Reservoirs for irrigation used by Pueblo 

Indians, remains of 178 

Rings, finger, of shell, from pueblo ruins in 

Arizona, discovery of 91 

Rio Tularosa , ruined pueblos on 180 

Robinson, Mr, aid given Dr Fewkes by 39 

Roof and floor construction at Homolobi, 

manner of 110 

Rough ware from Four-mile ruin, descrip- 
tion and figure of 141 

See also Pottery, surface finish of, rough. 

Russell, Frank, collections made by, in 

Arizona xxxix 

field work of ". . . xiii 

office work by, on architectural types 

of southern Arizona xxiii 

Rust, H. N., collection of stone objects from 

California, purchase of xxxix 

Sacrificial caves. See Graham mountains; 

Nan tacks. 
Sand fleas, plague of, a possible cause of 
abandonment of pueblos on 

Little Colorado river 22 

Sandia, inhabitants of, descent of 19 

visit to 18, 19 

San Jose\ Ariz., cinerary urn found near.. 181-18 

Pueblo pottery found near 173 

San Pedro river, field work along xii 

San Rafael valley, field work in xiii 

Santa Catalina range, field work in xiii 

Saucer from Four-mile ruin, decoration on. 142 

Savagery, social organization in xxiv-xxvii 

Saville. M. H., aid rendered by 253,299 

photographs of inscriptions from Cen- 
tral American monuments fur- 
nished by 253, 299, 301 

Saws or files of stone from pueblo ruins, 

description of 103, 159 

Seeds, occurrence of, in mortuary bowls 

at ruined pueblos in Arizona. . . 101 
Selenite plates, finding of, in pueblo ruins . Ill 

Seler, Eduard, on Cakchiquel calendar 279 

on Mayan chronologic tables. 245-246,257-258 
on significance of certain face numerals 

in Mayan inscriptions 226 

on significance of terminal dates of Ma- 
yan inscriptions 291-292 

Sharp, J. H., Indian portraits by, pur- 
chase of xxxix 

Shell ornaments from pueblo ruins, char- 
acter of 88-98, 1S7 

Shumopovi (Old), ruins of, bowl with snout 

from, cut showing Ill 


Shumopovi (Old), excavations at, stoppage 

of 112 

general features of 111-113 

pottery from, bird designs on 115-118 

decoration of 114-119 

gaming-cane design on, cut show- 
ing 118 

general features and form of 113 

rain-cloud design on, mention of . . . 156 
reptilian design on, description of. . 115 

figure showing 114 

specimens collected at, number of 18 

views of xlviii, xlix, l, li 

Shumway, Ariz., pueblo ruin near 164 

Sign language, reference to xxx 

Sikyatki, pottery from, decoration on, gen- 
eral character of 133, 153 

pottery from, features of 151, 153 

rain-cloud symbols on, mention of . 156 
Sitgreave, L., on features of Chevlon creek. 31 
ruined pueblos in- Arizona figured by . . 35 
Sitting Bull belt, obtained for National Mu- 
seum collection xxxix 

Skeletons from pueblo ruins, mention of. 128, 166 
Skeletons and mortuary objects from Four- 
mile ruin 139-140 

Skulls, human, from pueblo ruins, collec- 
tion of 34, 110, 1 34 

Sky-band, absence of, from specimens from 

Little Colorado ruins 74 

Slabs of stone. See Stone slabs. 

Slavery among primitive peoples, researches 

on xxiv-xxvii 

Slipper-shaped vessels from pueblo ruins, 

descriptions and figures of 69,181 

Snake (Plumed) on pottery from pueblo 

rnins, description of 115 

figure showing 114 

Snake dance, reference to 18 

Mishongnovi variant of, observation 

of 120 

Snowflake, Ariz., pueblo ruins near 121, 164 

Sociology, office work in xxi v-xxx 

Solomonville, Ariz., pueblo ruins near 171 

Sonora, Mexico, field work in ix 

Sonoyta valley, Mexico, field work in xiii 

Sophiology, office work in xxxiii-xxxvii 

Specimens collected from Arizona pueblo 
ruins, number and general 

character of 18 

Spherical stones from Pueblo Viejo ruins, 

description of 183 

Spider, importance of, in Pueblo mythology . 81 

figure of, on bowl from Homolobi 81, 82 

Spider woman, legends concerning 81, 82 

Spiral design on pottery from pueblo ruins, 

cuts showing 151, lxiv 

Springs, pueblosgenerally located near. 55 (note) 
Squash people, former inhabitants of Homo- 
lobi 24 

Stein, Robert, field observations by, on Es- 
kimo settlements xiv 

Steiner collection of stone implements, pur- 
chase of xxxix 

Stevenson, Mrs M. C, work of, on Zufii 

myths and ceremonies xxxvi 

PART 1] 



Stone from Four-mile ruin, used in belt 

frame, description of 159-160 

figure of 160 

Stone ax from Homolobi, figure showing .. 102 
Stone balls from Pueblo Viejo ruins, de- 
scription of 183 

Stone birds from Woodruff butte, descrip- « 

tion and figures of 135 

Stone fetish from Pueblo Viejo ruins, cut 

si lowing 186 

Stone hatchets from Pueblo Viejo ruins, 

collection of 183 

Stone implements from pueblo ruins, de- 
scriptions of 102-104, 

133, 159-160, 183-185 

figures showing 102, 182, 183, 184 

Rust collection of, purchase of xxxix 

Steiner collection of, purchase of xxxix 

Stone objects from shrine at Kintiel, men- 
tion of 127-128 

Stone slabs from pueblo ruins in Arizona, 

descriptions of 104-106, 

* 160-162, 185-186 

figures showing 161, 185, 186, xlvi, lxv 

perforations in 106, 160-162 

Stones, perforated, use of, by Pueblo In- 
dians 186 

See also Stone slabs. 
Stott ranch ruin, location and character of . . 168 

pottery from, character of 167 

Sulphur, fragment of, found at Chevlon ... Ill 
Sun, Temple of. at Palenque, face numerals 

at, value of 219-220 

Sun emblem on pottery from pueblo ruins, 

cut showing 150 

reference to 82 

Sun people. See Tawa people. 
Sunset, a Mormon town near site of Homo- 
lobi, cause of abandonment of. . 21 
Surface finish, classification of ware from 

pueblo ruins by 58-61, 179, 180 

Swanton, John R., work by xii, xxxii-xxxiii 

Swastika, wide range and diversity of, 

among American aborigines. . xv-xvi 

Swisshelm mountains, field work in xiii 

Symbolic types of art, study of xvii-xviii 

Symbolism of American aborigines, office 

researches relating to xv-xvi 

Symbols, numerical, of Mayan inscriptions. 

See Mayan. 
Symbols, quatern, wide range and diversity 

of xv-xvi 

Tanner's spring. Arizona, pueblo ruins near 127 
Tawa people, former inhabitants of Homo- 
lobi 24 

Tcipiya pueblo, location of 23 

Tciibkwitcalobi, name of, origin of :'>:{ 

Set also Chaves pass ruin. 

Technology, office work in xix-xxiii 

Tempe, people of, relations of, to people <.i 

Verde valley 34 

Templeof the Sun at Palenque, face numer- 
als at, values of 21'.) 

Tepoka Indian-, extinction of xi 

habitations of, discovery oi trace- of . .. xi 

Tesuki, visit to 18 

Thatcher, pueblo ruins near, mention of... 173 
Thomas, Cyrus, aid by, in linguistic classifi- 
cation xxxi-xxxii 

Mayan calendar systems, paper on ... 197-305 

characterization of paper by xliii-xliv 

work of, on hieroglyphic records xxxvi 

Thomas, Miss Jessie, work of, on translation 

of Diccionario de Motul xxxiii 

Tiguas, Mesa de las, site of Payiipki so 

called by Menchero 19 

Tikal, Central America, numeral system 

used at 254 

Time periods of Mayan chronology, terms 

designating 204 

Tonto valley, field work in xiii 

Tribal activities, definition of x 

Tribes, classification of x 

cyclopedia of native, work on xxxvii 

Pueblo, prehistoric migrations of, study 

of xviii-xix 

Troano codex, references to 252, 253 

Trumbull, J. H., Natick dictionary com- 
pleted by xxxi 

Tubes made of bone, from pueblo ruins, 

description of 94-95 

figure showing 95 

Tularosa, Rio, ruined pueblos on 180 

Turkey, use of, by Pueblo Indians 27-28 

Turkey Tank caves, Arizona, description of. 38-39 

view of iv 

Turquoise, Pueblo ornaments of, descrip- 
tions of 86 

Pueblo armlet inlaid with, figure show- 
ing 90 

Turritella shells, Pueblo articles made of. . . 91 
Turtle carapaces from pueblo ruins at 

Chevlon, description of 95 

disk of, cut showing 96 

Tusayan, map of 21 

visit to, reference to 120 

Tusayan pottery, characteristic color of . . . 192 
decorations on, superiority of 129 

Urns, cinerary, from Pueblo Viejo, descrip- 
tions of 173, 181-182 

plates showing LXVII, lxixc 

Vase, coiled, from Kintiel ruin, description 

a nd figure of 1 30 

Vases from pueblo ruins in Arizona, bird 

design-- on 75-76 

description of 61, 181 182, 189 

human figures on, description of 71 

views of (is, 7:;. 75, 

I'll, XX-XXII, xxx-xxxiii, XXXVIII, 
Set also Pottery. 
Vases in human effigy from pueblo ruins 

in Arizona, descriptions of... L89-192 

figures showing 189, l'.'i 

Verde valley, people of, relations of, to peo- 
ple of Casa Grande and Tempe. 34 
pottery from, similarity of, to that from 
chae, jmss and Gila-Salado 
basin 34 



[ETH. ANN. 22] 


Viejo. See Pueblo Viejo. 

Vigesimal system, use of, in Mayan chro- 
nology 217,234-236 

Virginia, field work in ix, xii 

Volcanic glass, arrowheads of, from the 

Naniacks 1*3 

Walpi, chiefs of, belonging to Patki clan ... 24 
symbolic types of art in, study of., xvii-xviii 

Flute ceremony at, reference to 18 

Water-house people, clans of 24 

traditional home of ancestors of .... 22, 24 

White ware, from ruined pueblos on Little 

Colorado river, reference to 61 

White and black ware from pueblo ruins. 59-60, 

180, XX, XXXII, xxxiv, 

xxxv, xxxvi, XLI 

White and green ware from pueblo ruins. . . 61, 


White and red ware from pueblo ruins, 

view showing xxxv 

White, black, and red ware from pueblo 

ruins 60-61, xxi-xxv, 


Winslow, Arizona, date of work near 17 

ruins between Hopi pueblos and, de- 
scriptions of 34-56 

ruins near, description of 25-34 

former inhabitants of 24-25 

identification of, as Homolobi 22 

location of 23 

Wisconsin, field work in ix 

Wood, H. S., editorial work of xxxviii 

Wood objects from Kintiel, mention of 133 

Woodruff butte, stone birds from, descrip- 
tion and figures of 135 

Wristlets of shell from pueblo ruins on Lit- 
tle Colorado river, discovery of. 90 
Written language, development of xxxi 

Xcalumkin, Yucatan, inscription at, discus- 
sit >n of 299-300 

inscription at, figure showing 253 

Yaki Indians, ceremonies of, account of . . xxx iv 
Yellow ware from pueblo ruins in Arizona, 

reference to 59 

views showing xxvu-xx vm 

Yellow, red, and brown ware, from pueblo 

ruins, plate showing xxv 

Yucatan, inscriptions and sculptures of, 

work on xxxvi 

inscription on tablet from 253 

Yucatec group of Mayan codices 252 

Yucca fiber, use of, by ancient Pueblo Indi- 
an < 97 

Zapeda, Sr Ramon, work of, as Mexican in- 
terpreter xii 

Zuck, Frank, aid given Dr Fewkes by 120-164 

cemetery of ancient Pinedale ruin dis- 
covered by - 166 

Zufii, visit by Dr Fewkes to J 18 

Zuni Indians, intermingling of, with Hopi 

'Indians, in ancient times 125 

legends of, concerning home of Katci- 

nas 144-145 

myths and ceremonies of, work on 

memoir of xxxvi 

quatern symbols used by x vi 

pottery of, inferiority of, as compared 

to that of Hopi Indians 129